CN117615017A - Calculation force request method, device and system - Google Patents

Abstract

This application provides a computing power request method, device and system, which belongs to the field of communication technology. In the solution provided by this application, after receiving the computing power request message, the first node corresponding to the first computing power service instance can determine whether to use the first computing power based on whether the first computing power service instance meets the service conditions. The service instance provides target computing power services to client devices. Since the first node can directly determine whether its corresponding first computing power service instance meets the service conditions and determine whether to provide the target computing power service, there is no need for routing nodes in the computing power network to sense the computing power of each computing power service instance. load, and select the matching computing service instance. As a result, the performance requirements for routing nodes are effectively reduced and the efficiency of providing computing power services is improved.

Description

Computing power request method, device and system

This application claims priority from the Chinese patent application with application number 202211004588.9 and the invention title "A computing power addressing method" submitted on August 22, 2022, the entire content of which is incorporated into this application by reference.

Technical field

The present application relates to the field of communication technology, and in particular to a computing power request method, device and system.

Background technique

Computing first network (CFN), referred to as computing network, is a network that provides computing services to client devices.

A computing power network generally includes one or more routing nodes and multiple service nodes used to provide computing power services. Among them, the computing power services provided by different service nodes can be the same or different. The routing node can sense the computing power load of each service node and generate a computing power routing table. The computing power routing table records the correspondence between the address of the service node and the computing power load of the service node. After the routing node receives the computing power request message sent by the client device, it can determine the matching service node from each service node based on the computing power load of each service node recorded in the computing power routing table, and use the computing power to The request message is routed to the matching service node.

However, in the above computing power request method, the routing node needs to sense the computing power load of a large number of service nodes, and determine the service node matching the client device based on the computing power load. As a result, the performance requirements for routing nodes are higher.

Contents of the invention

This application provides a computing power request method, device and system, which can solve the technical problem that routing nodes need to sense the computing power load of a large number of service nodes, resulting in high performance requirements for routing nodes.

In a first aspect, a computing power request method is provided, which is applied to a first node, the first node corresponds to a first computing power service instance, and the first computing power service instance is used to provide the target computing power service. The method includes: receiving a computing power request message from a client device, the computing power request message being used to request a target computing power service; and based on whether the first computing power service instance satisfies service conditions, determining whether the first computing power service instance is used by the first computing power service instance. The computing power service instance provides the target computing power service to the client device.

After receiving the computing power request message, the first node can directly determine whether its corresponding first computing power service instance meets the service conditions, and determine whether the first computing power service instance provides the target computing power service. As a result, routing nodes in the computing network no longer need to centrally sense the computing load of a large number of computing service instances and select matching computing service instances, thereby effectively reducing the performance requirements for routing nodes and improving the provision of computing power. Efficiency in service.

Based on the above analysis, it can be seen that the solution provided by this application can use the instance nodes corresponding to each computing service instance to determine the computing service instance used to provide the target computing service through distributed decision-making, thereby effectively alleviating the need to concentrate on routing nodes. pressure to release information when making decisions.

Optionally, the segment list of the computing power request message may include multiple segment identifiers; the multiple segment identifiers respectively indicate multiple computing power service instances, and the multiple computing power service instances are respectively used to provide the target computing power. service, the plurality of computing power service instances include a first computing power service instance.

By encapsulating a segment list including multiple segment identifiers in the computing power request message, the computing power request message can be forwarded to the instance nodes corresponding to each computing power service instance in sequence according to the forwarding path indicated by the segment list. Furthermore, each instance node that receives the computing power request message can determine whether the computing power service instance provides the target computing power service based on whether its corresponding computing power service instance meets the service conditions. That is, the instance node corresponding to each computing power service instance can determine the computing power service instance used to provide the target computing power service through distributed decision-making.

Optionally, based on whether the first computing power service instance satisfies the service condition, the process of determining whether the first computing power service instance provides the target computing power service to the client device may include: based on the first computing power service instance satisfying the service condition Conditions, determine that the first computing power service instance provides the target computing power service to the client device, and sends a computing power response message to the client device; or, based on the first computing power service instance not meeting the service conditions, determine The first computing power service instance provides the target computing power service to the client device, and forwards the computing power request message according to the segment list.

After the first node sends the computing power response message, the client device can continue to send the computing power task message to the first node based on the response message to obtain the target computing power service. After the first node forwards the computing power request message according to the segment list, the instance node that subsequently receives the computing power request message can continue to judge that its corresponding computing power service instance meets the service conditions, and then determine whether its corresponding computing power service instance satisfies the service conditions. Power service instances provide target computing power services to client devices.

Optionally, the destination address field of the computing power request message may include a first segment identifier used to indicate the first computing power service instance, and the plurality of segment identifiers include the first segment identifier. Correspondingly, the process of forwarding the computing power request message according to the segment list may include: updating the value of the destination address field of the computing power request message to the second segment identifier in the segment list, and forwarding the computing power request message to Forwarded to the second node. The second node corresponds to the second computing power service instance indicated by the second segment identifier, and the second segment identifier is the next segment identifier adjacent to the first segment identifier in the segment list.

In the solution provided by this application, the computing power request message encapsulated with SRH is an SRv6 message. After the first node determines that the first computing power service instance does not meet the service conditions, it can forward the computing power request message based on the SRv6 message forwarding mechanism. This enables the flexible application of SRv6 technology in computing power networks.

Optionally, before forwarding the computing power request message according to the segment list, the method may further include: encapsulating the computing power load of the first computing power service instance in the computing power request message.

By encapsulating the computing power load of the computing power service instance in the computing power request message, the instance node corresponding to the last segment identifier in the segment list can obtain the computing power load of each computing power service instance. Furthermore, when none of the computing power service instances indicated by each segment identifier in the segment list meets the service conditions, the instance node corresponding to the last segment identifier can be selected to provide the service to the customer based on the computing power load of each computing power service instance. The terminal device provides the target computing power service instance of the target computing power service.

Optionally, the destination address field of the computing power request message may include a first segment identifier used to indicate the first computing power service instance, and the plurality of segment identifiers include the first segment identifier. Correspondingly, the process of forwarding the computing power request message according to the segment list may include: based on the first segment identifier being the last segment identifier in the segment list, updating the value of the destination address field of the computing power request message to segment The third segment identifier in the list gets the service indication; then, the service indication is forwarded to the third node. The third node corresponds to the third computing power service instance indicated by the third segment identifier, and the service indication is used to instruct the third computing power service instance to provide the target computing power service to the client device.

In the solution provided by this application, when the computing power service instance indicated by the last segment identifier in the segment list also does not meet the service conditions, the first node can forcibly instruct the third computing power service instance to provide the target computing power to the client device. Serve. This ensures that the client device can effectively obtain the target computing power service.

Optionally, the third computing power service instance may be randomly selected from multiple computing power service instances. Alternatively, the third computing power service instance may satisfy at least one of the following conditions: the computing power load of the third computing power service instance is less than the first load threshold; the third node corresponding to the third computing power service instance is not the same as the first The network performance between routing nodes meets the first communication condition. Wherein, the first routing node is a routing node connected to the client device.

Based on the above conditions, it can be ensured that the computing power service instance with smaller computing power load and/or better network performance can provide the target computing power service to the client device.

Optionally, based on whether the first computing power service instance satisfies the service condition, the process of determining whether the first computing power service instance provides the target computing power service to the client device may include: based on the first segment identification as a segment list The last segment identifier in determines that the first computing power service instance provides the target computing power service to the client device and sends a computing power response message to the client device.

In the solution provided by this application, when the first segment identifier is the last segment identifier in the segment list, the first node can determine that none of the computing power service instances indicated by other segment identifiers meet the service conditions. Therefore, the first node can directly determine that its corresponding first computing power service instance provides the target computing power service to the client device.

Optionally, the computing power response message sent by the first node may include the first segment identification. Therefore, it is convenient for the client device to directly send the computing power task message to the first node based on the first segment of the identifier to obtain the target computing power service.

Optionally, the source address field or extension header of the computing power response message includes the first segment identifier. The extended message header may be SRH.

Optionally, the segment list may also include the identification of the target computing power service. For example, the last segment identification field in the segment list may include the identification of the target computing power service.

Based on this, it is convenient for the routing node or service node in the computing power network to determine the source address and destination of the computing power request message when the destination address field in the computing power request message is updated to identify the target computing power service. The address can pass the verification during verification to avoid discarding the computing power request message due to verification failure.

Optionally, the service condition may include: the computing power load of the first computing power service instance is less than the second load threshold. This ensures that the computing power service instance with a smaller computing power load provides the target computing power service to the client device.

Optionally, the service condition may also include: network performance between the first routing node connected to the client device and the first node meets the second communication condition. This ensures that the computing power service instance with better network performance provides the target computing power service to the client device.

Optionally, the first node may be a service node to which the first computing power service instance belongs; or, the first node may be a second routing node directly connected to the service node to which the first computing power service instance belongs.

Wherein, if the first node is a service node, the service node can determine whether the first computing power service instance satisfies the service condition based on the computing power load of the first computing power service instance deployed by it. As a result, routing nodes in the computing power network no longer need to sense the computing power load of the computing power service instance and only need to forward messages, thus effectively reducing the control surface pressure of routing nodes.

If the first node is a second routing node, the second routing node can only sense the computing load of the computing service instance deployed in the service node to which it is directly connected, and the service node does not need to parse the SRH in the message and calculate the computing power load according to the SRH in the message. Segment list forwards the message. As a result, it can be effectively compatible with the functions of traditional service nodes while effectively reducing the control surface pressure of the routing node.

In a second aspect, a computing power request method is provided. The method includes: generating a computing power request message. The segment list of the computing power request message includes multiple segment identifiers. The multiple segment identifiers respectively indicate multiple computing power requests. A computing power service instance, each of the multiple computing power service instances is used to provide a target computing power service. Afterwards, the computing power request message is forwarded according to the segment list, so that at least one computing power service instance among the plurality of computing power service instances that meets the service condition provides the target computing power service to the client device.

By encapsulating the segment list in the computing power request message, the computing power request message can be forwarded to the instance nodes corresponding to each computing power service instance along the forwarding path indicated by the segment list. Furthermore, each instance node that receives the computing power request message can determine the computing power service instance used to provide the target computing power service through distributed decision-making. This effectively reduces the pressure of information release when centralized decision-making by routing nodes.

Optionally, the computing power request method can be applied to the client device. Correspondingly, the process of generating the computing power request message may include: based on the target computing power service requested by the client device, encapsulating multiple segment identifiers corresponding to the target computing power service into the computing power request message.

By generating a computing power request message through the client device, the routing node connected to the client device no longer needs to sense and obtain multiple segment identifiers corresponding to the target computing power service, thus effectively simplifying the operation of the routing node.

Optionally, before generating the computing power request message, the method may also include: sending a domain name request for the target computing power service to a domain name system (DNS) server, and receiving a domain name request related to the target computing power service sent by the DNS server. Multiple segment identifiers corresponding to force services.

Optionally, the computing power request method can be applied to a routing node connected to the client device. Correspondingly, the process of generating a computing power request message may include: receiving an initial request message sent by the client device, where the initial request message includes the identification of the target computing power service; encapsulating the information related to the target computing power service in the initial request message. Multiple segment identifiers corresponding to the identifier of the computing power service, and obtain the computing power request message.

By generating the computing power request message through the routing node, the client device no longer needs to sense and obtain multiple segment identifiers corresponding to the target computing power service, thus effectively simplifying the operation of the client device.

Optionally, before the routing node generates the computing power request message, the method may also include: receiving multiple segment identifiers corresponding to the identifier of the target computing power service advertised by other routing nodes. Other routing nodes may advertise the multiple segment identifiers through a control plane protocol (such as BGP).

Optionally, the process of the routing node generating the computing power request message may also include: encapsulating the network performance parameters between the routing node and the instance node corresponding to the computing power service instance indicated by the multiple segment identifiers into the computing power request. message.

By encapsulating network performance parameters in the computing power request message, the instance node that receives the computing power request message can jointly determine whether the computing power service instance meets the service conditions based on the computing power load and network performance of the computing power service instance.

Optionally, the computing power request message may include a segment routing header (SRH), and an extended type-length-value (TLV) field in the SRH includes the network performance parameter.

Optionally, before encapsulating the multiple segment identifiers corresponding to the identifier of the target computing power service in the initial request message, the method may also include: selecting from multiple alternative segments corresponding to the identifier of the target computing power service. The multiple segment identifiers that meet the encapsulation conditions are filtered out from the identifiers. The encapsulation conditions may include: the network performance between the routing node and the instance node corresponding to the computing power service instance indicated by the segment identifier meets the communication conditions.

Since the first routing node can filter out instance nodes with poor network performance in advance, it can ensure that only instance nodes with better network performance determine whether their corresponding computing power service instance provides the target computing power service. This not only ensures the efficiency of determining the computing power service instance that meets the service conditions, but also ensures the reliability of providing the target computing power service to the client device.

Optionally, the multiple segment identifiers may be arranged in order from high to low network performance. Based on this, it can be ensured that the instance node with better network performance between the routing node and the routing node can receive the computing power request message sooner, thereby ensuring that the instance node with better network performance is prioritized to detect whether its corresponding computing power is used Service instances provide target computing power services.

Optionally, after forwarding the computing power request message according to the segment list, the method may also include: receiving a computing power response message sent by the target node, where the computing power response message includes the target computing power corresponding to the target node. The target segment ID of the service instance that belongs to the multiple segment IDs. Afterwards, a computing power task message is sent to the target node, and the destination address field of the computing power task message includes the target segment identifier.

In the solution provided by this application, the client device can send a computing power task message to the target node, and the target segment identifier in the computing power task message can be encapsulated by the client device or connected by the client device. routing node encapsulation.

Optionally, the source address field or extension header of the computing power response message includes the target segment identifier.

Optionally, the identifier of the target computing power service may be a common prefix of the multiple segment identifiers. Correspondingly, before sending the computing power task message to the target node, the method may also include: based on the prefix of the target segment identifier matching the identifier of the target computing power service, recording the corresponding relationship between the target computing power service and the target segment identifier. Thus, the target segment identifier can be verified to ensure the reliability of the recorded correspondence.

Optionally, the segment list may also include the identification of the target computing power service.

In the third aspect, a computing power request method is provided, which can be applied to the computing power request system. The computing power requesting system includes a requesting node and multiple instance nodes corresponding to multiple computing power service instances, where each computing power service instance in the multiple computing power service instances is used to provide a target computing power service. The method includes: requesting a node to generate a computing power request message, and the segment list of the computing power request message includes multiple segment identifiers, and the multiple segment identifiers respectively indicate multiple computing power service instances. Afterwards, the requesting node forwards the computing power request message according to the segment list. The first node among the plurality of instance nodes receives the computing power request message, and the first node corresponds to the first computing power service instance among the plurality of computing power service instances. After that, the first node based on the first computing power service Whether the instance satisfies the service conditions and determines whether the first computing power service instance provides the target computing power service to the client device.

Optionally, the first node determines whether the first computing power service instance provides the target computing power service to the client device based on whether the computing power service instance satisfies the service condition. The process may include: the first node determines whether the first computing power service instance satisfies the service condition. The service instance meets the service conditions, determines that the first computing power service instance provides the target computing power service to the client device, and sends a computing power response message to the client device; or, the first node based on the first computing power service instance If the service conditions are not met, it is determined that the first computing power service instance does not provide the target computing power service to the client device, and the computing power request message is forwarded to the second node among the multiple instance nodes according to the segment list.

Optionally, the computing power response message includes a first segment identifier used to indicate the first computing power service instance; the method may also include: requesting the node to send a computing power task message to the first node, the computing power The destination address field of the task message includes the first segment identifier.

Optionally, the requesting node is a client device; before the requesting node generates the computing power request message, the method may also include: the client device sends a domain name request for the target computing power service to the DNS server; the client device receives Multiple segment identifiers corresponding to the target computing power service sent by the DNS server.

Optionally, the requesting node is a routing node connected to the client device; before the requesting node generates the computing power request message, the method may also include: the routing node receives an initial request message sent by the client device, and the The initial request message includes the identifier of the target computing service; the routing node determines multiple segment identifiers corresponding to the identifier of the target computing service.

In the fourth aspect, a computing power request method is provided, which is applied to the tail instance node. The method includes: receiving a computing power request message, the segment list of the computing power request message includes multiple segment identifiers, the multiple segment identifiers respectively indicate multiple computing power service instances, wherein each computing power service instance is used for Provide the target computing power service. The tail instance node corresponds to the computing power service instance indicated by the last segment identifier in the segment list. Afterwards, the tail instance node can send a service instruction to the instance node corresponding to the target computing power service instance among the multiple computing power service instances, so that the target computing power service instance provides the target computing power service to the client device.

Optionally, the computing power request message may also include the computing power load of other computing power service instances among the multiple computing power service instances except the computing power service instance corresponding to the last instance node. The tail instance node can determine the target computing power service instance from the multiple computing power service instances based on the computing power load of the multiple computing power service instances.

Optionally, the target computing power service instance may satisfy: the computing power load of the target computing power service instance is less than the first load threshold.

Optionally, the target computing power service instance may also satisfy: the network performance between the instance node corresponding to the target computing power service instance and the first routing node meets the first communication condition. Wherein, the first routing node is a routing node connected to the client device.

Optionally, the target computing power service instance may be randomly selected by the tail instance node from multiple computing power service instances.

In the fifth aspect, a computing power request method is provided, which is applied to the instance node corresponding to the computing power service instance. The instance node can be the service node to which the computing power service instance belongs, or it can be a route directly connected to the service node. node. The method includes: receiving a computing power request message from a client device. The segment list of the computing power request message includes multiple segment identifiers. The multiple segment identifiers respectively indicate multiple computing power service instances, wherein each computing power request message includes multiple segment identifiers. Power service instances are used to provide target computing power services. Afterwards, the computing power load of the corresponding computing power service instance is encapsulated in the computing power request message, and the computing power request message is forwarded according to the segment list.

Optionally, before forwarding the computing power request message according to the segment list, the method may also include: encapsulating the network performance parameters between the computing power request message and the first routing node in the computing power request message. Wherein, the first routing node is a routing node connected to the client device.

A sixth aspect provides a first node, which may be a service node or a routing node directly connected to the service node. Moreover, the first node includes at least one module, and the at least one module can be used to implement the computing power request method provided in the above-mentioned first aspect, fourth aspect or fifth aspect.

A seventh aspect provides a request node, which may be a client device or a routing node directly connected to the client device. Furthermore, the request node includes at least one module, and the at least one module can be used to implement the computing power request method provided in the second aspect.

In an eighth aspect, a computing power requesting device is provided. The computing power requesting device includes: a memory, a processor, and a computer program stored in the memory and capable of running on the processor. When the processor executes the computer program Implement the computing power request method provided by any of the above aspects.

In a ninth aspect, a computer-readable storage medium is provided. Instructions are stored in the computer-readable storage medium. When the instructions are run on a processor, they cause the processor to execute the computing power request method provided in any of the above aspects. .

A tenth aspect provides a computer program product containing instructions that are executed by a processor to implement the computing power request method provided in any of the above aspects.

An eleventh aspect provides a chip that can be used to implement the computing power request method provided in any of the above aspects.

In a twelfth aspect, a computing power requesting system is provided. The computing power requesting system includes: multiple instance nodes corresponding to multiple computing power service instances, each computing power service instance in the multiple computing power service instances. Used to provide target computing power services. Furthermore, at least one instance node among the plurality of instance nodes is used to implement the computing power request method provided in the above-mentioned first aspect, fourth aspect or fifth aspect.

Optionally, each instance node among the plurality of instance nodes may be a service node to which its corresponding computing power service instance belongs; or, each instance node among the plurality of instance nodes may be a service node to which its corresponding computing power service instance belongs. The routing node directly connected to the service node to which the force service instance belongs.

Optionally, the system may also include: a request node, the request node being used to implement the computing power request method provided in the second aspect above. The requesting node may be a client device, or a routing node connected to the client device.

To sum up, this application provides a computing power request method, device and system. In the solution provided by this application, after receiving the computing power request message, the first node corresponding to the first computing power service instance can determine whether to be served by the first computing power based on whether the first computing power service instance meets the service conditions. Instances provide target computing power services to client devices. Since the first node can directly determine whether its corresponding first computing power service instance meets the service conditions and determine whether to provide the target computing power service, there is no need for routing nodes in the computing power network to centrally sense the computing power of a large number of computing power service instances. load, and select a matching computing service instance. That is to say, the method provided by this application can determine the computing service instance used to provide the target computing service through distributed decision-making, thereby effectively reducing the pressure of information release when centralized decision-making by routing nodes. Furthermore, it effectively reduces the performance requirements for routing nodes and improves the efficiency of providing computing power services.

Description of drawings

Figure 1 is a schematic structural diagram of a computing power network provided by an embodiment of the present application;

Figure 2 is a schematic structural diagram of another computing power network provided by an embodiment of the present application;

Figure 3 is a schematic structural diagram of an SRv6 message provided by an embodiment of the present application;

Figure 4 is a schematic structural diagram of a SID provided by an embodiment of the present application;

Figure 5 is a schematic diagram of an SRv6 message forwarding process provided by an embodiment of the present application;

Figure 6 is a flow chart of a computing power request method provided by an embodiment of the present application;

Figure 7 is a flow chart of another computing power request method provided by an embodiment of the present application;

Figure 8 is a flow chart of yet another computing power request method provided by an embodiment of the present application;

Figure 9 is a schematic diagram of forwarding a computing power request message provided by an embodiment of this application;

Figure 10 is a flow chart of yet another computing power request method provided by an embodiment of the present application;

Figure 11 is a schematic diagram of forwarding another computing power request message provided by the embodiment of this application;

Figure 12 is a flow chart of yet another computing power request method provided by an embodiment of the present application;

Figure 13 is a schematic structural diagram of a first node provided by an embodiment of the present application;

Figure 14 is a schematic structural diagram of a request node provided by an embodiment of the present application;

Figure 15 is a schematic structural diagram of a computing power request device provided by an embodiment of the present application;

Figure 16 is a schematic structural diagram of another computing power requesting device provided by an embodiment of the present application.

Detailed ways

The computing power request method, device and system provided by the embodiments of the present application will be introduced in detail below with reference to the accompanying drawings.

Artificial intelligence technology brings about the transformation of the information society into a smart society. A typical feature of the smart society is the deep integration of the physical world and the digital world. In the future, the digital world will interact with the real world through sensors and actuators provided by technologies such as the Internet of Things and augmented reality. . The interaction mode has changed from the interaction between people in traditional society to the interaction between data and computing power. The primary demand of users and applications using the Internet may no longer be to obtain information, but to obtain computing resources and functions. The core purpose is to obtain The result of the calculation.

Based on the above needs, the computing power network emerged as the times require. The technical components of the computing network mainly include computing network collaborative scheduling on the control plane, network fusion sensing on the data plane, and computing resource orchestration on the management and service plane. The overall architecture of the computing power network system should have the ability to uniformly manage underlying computing resources, storage resources and network resources, and be able to measure underlying infrastructure resources with unified standards, abstract them into information elements and load them into network messages. Share over the network.

The computing power network can realize edge computing networking and edge-to-edge collaboration, using the multi-copy feature of services to achieve nearby access for users and load balancing of services, and adapt to the dynamics of services. Moreover, the computing power network is based on service identity (service ID) addressing and publishes computing power status and network status to the network as routing information. The network routes the computing power request message to the most appropriate computing node to achieve the best user experience. Optimal, optimal computing resource utilization and optimal network efficiency.

Figure 1 is a schematic structural diagram of a computing power network provided by an embodiment of the present application. As shown in Figure 1, the computing power network may include multiple routing nodes 01 and multiple service nodes 02. Communication connections are established between the multiple routing nodes 01 , and a communication connection is established between each service node 02 and one routing node 01 . Each routing node 01 may be a device with a packet forwarding function such as a router or a switch, which may also be called a computing power routing device or a CFN node.

A computing power service instance is deployed in each service node 02, and the computing power service instance can be used to provide computing power services, that is, to provide computing resources, functions or services. Each service node 02 can be a single physical server, a server cluster, or a cloud server. The service node 02 may also be called a computing node, or a multi-access edge computing (MEC) node. Referring to Figure 1, it can be seen that the multiple routing nodes 01 may be located at the CFN layer, and the CFN layer may be underlayed by the Internet protocol (Internet protocol, IP) layer.

Continuing to refer to Figure 1, the routing node 01 can also be connected to the client device 03, and is used to route the computing power request message sent by the client device 03 to the matching service node 02, so that the matching service node 02 is the client Device 03 provides computing power services. The client device 03 may also be called a user device, which may be a residential gateway (RGW), a mobile phone, a laptop or a desktop computer, or other devices.

It can be understood that the routing node 01 directly connected to the client device 03 may also be called an ingress node, and the routing node 01 directly connected to the matching service node 02 may also be called an egress node. The plurality of service nodes 02 can be divided into edge nodes, regional nodes, center nodes, etc. according to different deployment locations. Moreover, the performance of service nodes 02 in different deployment locations may be different. For example, the performance of the central node may be better than that of the edge node.

In related technology, after routing node 01 receives a computing power request message sent by a certain client device 03, it can determine the matching service node 02 based on the pre-stored computing power routing table, and generate a flow table for the client device 03. And forward the computing power request message of client device 03 based on the flow table. Referring to Figure 2, assume that CFN node 1 in the computing power network is connected to the client device, CFN node 2 is connected to service node 1, and CFN node 3 is connected to service node 2. The IP address of service node 1 is IPA, the IP address of service node 2 is IPB, and the computing service instances deployed in the two service nodes are used to provide the same computing service. Then the process of CFN node 1 forwarding the computing power request message of the client device may include:

Step 1: CFN node 2 and CFN node 3 respectively notify CFN node 1 of the association between the computing power load and the network location through the border gateway protocol (BGP) protocol. Among them, the computing power load can also be called the computing power status, and the network location can be the IP address of the service node. For example, the association relationship advertised by CFN node 2 includes the identifier S1 of the computing power service provided by service node 1, the computing power load of service node 1, and the IP address of service node 1. The association relationship advertised by CFN node 3 includes the identifier S1 of the computing power service provided by service node 2, the computing power load of service node 2, and the IP address of service node 2.

Among them, the identity of the computing power service can also be called a service identity. The computing power load of each service node can be characterized by at least one of the following parameters: processor usage, memory usage, number of online connections, network delay and integration overhead (metric), etc. The processor usage may include a central processing unit (CPU) usage, and/or a graphics processing unit (GPU) usage, etc.

Step 2: CFN node 1 establishes a computing power routing table based on the received association relationship. The computing power routing table can maintain different service nodes corresponding to the same computing power service, as well as the computing power loads of different service nodes. For example, the computing power routing table established by CFN node 1 can be as shown in Table 1. Among them, the computing power service with the service identifier S1 corresponds to two service nodes, and the IP addresses of the two service nodes are IPA and IPB respectively. The computing power load of each service node can include CPU usage, GPU usage, number of online connections, network delay and integration overhead.

Table 1

Step 3: The client device that needs to access remote network services sends a computing power request message to CFN node 1 (that is, the head-end CFN node). The destination address (DA) of the computing power request message is the identification of the requested computing power service. For example, referring to Figure 2, the destination address of the computing power request message is S1.

Step 4: CFN node 1 determines the matching service node based on the computing power routing table and the computing power selection strategy (for example, giving priority to lightly loaded service nodes), and updates the destination address of the computing power request message to the IP address of the service node. And forward the computing power request message to the service node. Furthermore, CFN node 1 can establish a flow table from the client device to the service node to guide subsequent packet forwarding. Among them, the flow table has an aging mechanism. After aging, the CFN node 1 needs to reselect matching service nodes based on the computing power load of each service node.

For example, assuming that the matching service node determined by CFN node 1 is service node 1, CFN node 1 can update the destination address of the computing power request message to IPA and forward it to service node 1. Moreover, CFN node 1 can establish a flow table as shown in Table 2. The flow table includes the five-tuple information of the client device, the matching IP address IPA of service node 1, the next hop address, and the outgoing interface. Among them, the five-tuple information includes source IP, source port, service identifier, destination port, and protocol type. The next hop address is the IP address of CFN node 2 connected to the service node 1: IP1. The outbound interface is an SRv6 tunnel.

Table 2

Quintuple service node Next hop Outbound interface Source IP, source port, service identifier, destination port, protocol type IPA IP1 SRv6 tunnel

Step 5: CFN node 1 forwards the computing power request message according to the next hop corresponding to the IP address of the service node in the flow table. For example, referring to Table 2, CFN node 1 can tunnel-encapsulate and forward the computing power request message.

Step 6: CFN node 2 decapsulates the tunnel and routes the computing power request message to the corresponding service node 1 according to the destination address of the computing power request message. Among them, CFN node 2 is the tail node of the transmission network.

Step 7: After CFN node 1 subsequently receives the computing power task message sent by the client device, it can directly forward the computing power task message according to the flow table shown in Table 2, without performing the above step 4.

However, the above service request method requires the CFN node to sense and maintain the computing power load of a large number of service nodes, resulting in greater pressure on the control plane of the CFN node. Moreover, for each client device and each service flow, the CFN node needs to maintain a corresponding flow table, resulting in a large number of table entries to be maintained in the CFN node and a heavy burden. In addition, when network flapping causes the flow table to be lost, the CFN node needs to reselect a matching service node. If the service nodes selected before and after the CFN node are inconsistent, the unfinished tasks in the previous service node need to be restarted by the reselected service node, which wastes computing resources.

The embodiment of this application provides a computing power request method. After receiving the computing power request message, the instance node corresponding to the computing power service instance (such as a service node or a routing node directly connected to the service node) directly determines its corresponding Whether the computing power service instance meets the service conditions, and then determines whether the computing power service instance provides computing power services to the client device. That is to say, the method provided by the embodiments of the present application enables each instance node to determine the computing power service instance used to provide computing power services to the client device through distributed decision-making. Compared with the centralized decision-making method of routing nodes, this distributed decision-making method does not require routing nodes to sense the computing power load of computing service instances deployed in a large number of service nodes, and does not require routing nodes to maintain a large number of flow tables. As a result, the performance requirements for routing nodes are effectively reduced and the efficiency of providing computing power services is improved.

Moreover, in this embodiment of the present application, the computing power request message may be an SRv6 message with an SRH inserted, and the segment list of the SRH includes segment identifiers of multiple computing power service instances used to provide the same computing power service. That is to say, the solution provided by the embodiments of this application can be combined with SRv6 technology to realize the forwarding of computing power request messages. The following describes the forwarding mechanism of SRv6 messages. Figure 3 is a schematic structural diagram of an SRv6 message provided by an embodiment of the present application. As shown in Figure 3, the SRv6 message includes: IPv6 header, SRH and payload. Among them, the IPv6 header includes the following fields: version (version), traffic class (TC), flow label (flow label), payload length (payload length), next header (nextheader), hop limit (hop limit) ), source address (source address, SA) and DA.

SRH includes the following fields: next header, header extension length (Hdr ExtLen), routing header type (routing type), SL, last entry, flags, tag, segment List and optional TLV. Among them, the segment list includes segment list [0] to segment list [n], a total of n+1 identification fields (n can be an integer greater than or equal to 1), and each identification field is used to carry a node in the message forwarding path. SID, which can be the IPv6 address of the node. Please refer to Table 3 for the length and meaning of each field in the SRH.

table 3

Figure 4 is a schematic structural diagram of an SID provided by an embodiment of the present application. As shown in Figure 4, each SID may include a locator field, a function field and an arguments field. Among them, the locator field is used to identify a certain network domain and a certain node in the network domain. The function field is used to identify the action performed on the corresponding node, also called behavior. The parameter field is used to carry the parameters required to perform the action.

During the packet forwarding process, a node (such as a routing node) may update the value of the DA field of the packet based on the SID indicated by the SL field. The content in the SRH is equivalent to a computer program. Segment list [0] to segment list [n] are equivalent to multiple instructions in the computer program, and the first instruction to be executed in the computer program is segment list [n] . The SL field in SRH is equivalent to the pointer of the computer program, which always points to the currently executing instruction. Before the message is forwarded, the value of the SL field can be initialized to n. After each instruction is executed, the value of the SL field can be decremented by 1 (that is, SL-- is executed), so that the SL field points to the next instruction to be executed. Through the above method, end-to-end forwarding of messages can be achieved.

For example, referring to Figure 5, assume that the forwarding path of the SRv6 message is: Node A→Node B→Node C→Node D→Node I. Among them, node A, node B, node D and node I all support SRv6 technology, while node C does not support SRv6 technology. Then node A, as the ingress node, can insert SRH into the SRv6 message. The SRH is encapsulated with the segment list <I::, D::, B::>. Among them, I:: is the SID of node I, D:: is the SID of node D, and B:: is the SID of node B. Before forwarding the message, node A can initialize the value of the SL field to 2. Then, node A can copy the SID in the segment list indicated by the SL field, that is, B:: in segment list [2], to the DA field in the IPv6 header. Afterwards, node A can forward the SRv6 packet to node B.

After receiving the SRv6 message, Node B can search the local SID table based on the value of the DA field in the IPv6 header. Since the value of the DA field can hit the local SID table of node B, node B can decrement the value of the SL field by 1 and add the SID in the segment list indicated by the updated SL field, that is, the SID in the segment list [1] D::Copied to the DA field in the IPv6 header. Afterwards, node B can forward the SRv6 packet to node C.

After node C receives the SRv6 message, since node C does not support SRv6 technology, it does not need to process SRH. Instead, it can directly query the routing table based on the DA in the IPv6 header and forward the SRv6 message to node D.

After receiving the SRv6 message, node D can search the local SID table based on the DA in the IPv6 header. Since the DA can hit the local SID table of node D, node D can decrement the value of the SL field by 1 and add the SID in the segment list indicated by the updated SL field, that is, I:: in segment list [0]. Copied to the DA in the IPv6 header. Since the value of the updated SL field is 0, node D can eject the SRH and forward the SRv6 message of the ejected SRH to node I.

The computing power request method provided by the embodiment of this application is introduced below. This method can be applied to a first node in a computing power network such as that shown in Figure 1. The first node corresponds to a first computing power service instance, and the first computing power service instance is used to provide a target computing power service. For example, the first node may be a service node where the first computing power service instance is deployed, or may be a routing node directly connected to the service node. Referring to Figure 6, the method includes:

Step 101: Receive a computing power request message from the client device. The computing power request message is used to request the target computing power service.

In this embodiment of the present application, the first node can receive a computing power request message from a client device, and the computing power request message is used to request a target computing power service. The computing power request message may be generated by the client device and sent to the first node. Alternatively, the computing power request message may be generated by a routing node directly connected to the client device and forwarded to the first node.

Step 102: Based on whether the first computing power service instance satisfies the service conditions, determine whether the first computing power service instance provides the target computing power service to the client device.

After receiving the computing power request message, the first node can detect whether the first computing power service instance meets the service conditions, and can determine whether the first computing power service instance provides the target computing power service to the client device based on the detection result. . For example, if the first node detects that the first computing power service instance meets the service conditions, it can determine that the first computing power service instance provides the target computing power service to the client device, and can send a computing power response message to the client device. arts. If the first node detects that the first computing power service instance does not meet the service conditions, it can determine that the first computing power service instance does not provide the target computing power service to the client device, and can continue to forward the computing power request to other instance nodes. message.

It can be understood that the segment list of the computing power request message may include multiple segment identifiers (SIDs), and the multiple segment identifiers respectively indicate multiple computing power service instances. That is to say, each segment identifier in the multiple segment identifiers can indicate a computing power service instance, that is, it can identify the unique position of a computing power service instance in the computing power network, and the computing power service instances indicated by different segment identifiers different. Moreover, each computing power service instance among the plurality of computing power service instances is used to provide a target computing power service. The destination address field of the computing power request message may include a first segment identifier, and the first segment identifier belongs to the multiple segment identifiers.

After receiving the computing power request message, the first node may first detect whether the first segment identifier matches the segment identifier of the first computing power service instance. If the first segment identifier matches the segment identifier of the first computing power service instance, the first node may continue to detect whether the first computing power service instance meets the service conditions. If the first segment identifier does not match the segment identifier of the first computing power service instance, the first node can forward the computing power request message based on the first segment identifier without detecting whether the first computing power service instance meets the service conditions. .

To sum up, the embodiments of this application provide a computing power request method. After the first node corresponding to the first computing power service instance receives the computing power request message, it can based on whether the first computing power service instance satisfies the service requirement. condition to determine whether the first computing power service instance provides the target computing power service to the client device. Since the first node can directly determine whether its corresponding first computing power service instance meets the service conditions and determine whether to provide the target computing power service, there is no need for routing nodes in the computing power network to centrally sense the computing power of a large number of computing power service instances. load, and select a matching computing service instance. That is to say, the method provided by the embodiments of the present application can determine the computing power service instance used to provide the target computing power service through distributed decision-making, thereby effectively reducing the pressure of information release when centralized decision-making by routing nodes. Furthermore, it effectively reduces the performance requirements for routing nodes and improves the efficiency of providing computing power services.

The embodiment of the present application provides another computing power request method, which can be applied to requesting nodes in the computing power network such as shown in Figure 1. The requesting node may be a client device, or may be a routing node directly connected to the client device. Referring to Figure 7, the method includes:

Step 201: Generate a computing power request message. The segment list of the computing power request message includes multiple segment identifiers.

In this embodiment of the present application, the requesting node may determine multiple segment identifiers corresponding to the target computing power service based on the requested target computing power service. Afterwards, the requesting node can encapsulate the SRH in the computing power request message, and the segment list of the SRH includes the multiple segment identifiers. The multiple segment identifiers respectively indicate multiple computing power service instances, and each of the multiple computing power service instances is used to provide a target computing power service.

Step 202: Forward the computing power request message according to the segment list.

After the requesting node generates the computing power request message, it can forward the computing power request message to the routing node in the computing power network, so that at least one computing power among the computing power service instances indicated by the multiple segment identifiers meets the service conditions. The service instance provides target computing power services to client devices. Among them, a computing power service instance indicated by each segment identifier corresponds to an instance node. The instance node can be a service node to which the computing power service instance belongs, or a routing node directly connected to the service node. After each instance node receives the computing power request message, it can determine whether its corresponding computing power service instance provides the target computing power service to the client device based on whether its corresponding computing power service instance meets the service conditions.

To sum up, embodiments of this application provide a computing power request method. The computing power request message generated by the requesting node includes a segment list, and the computing power request message can be forwarded based on the segment list. Since each segment identifier in the segment list indicates a computing power service instance used to provide the target computing power service, the computing power request message can be forwarded to the corresponding computing power service instance along the forwarding path indicated by the segment list. Instance node. Furthermore, the instance node that receives the computing power request message can determine whether the computing power service instance provides the target computing power service to the client device based on whether its corresponding computing power service instance meets the service conditions. As a result, routing nodes in the computing network no longer need to centrally sense the computing load of a large number of computing service instances and select matching computing service instances, thereby effectively reducing the performance requirements for routing nodes and improving the provision of computing power. Efficiency in service.

Figure 8 is a flow chart of another computing power request method provided by an embodiment of the present application. In this method, the requesting node is the first routing node directly connected to the client device, and the first node is used as the first computing power service. The second routing node directly connected to the service node to which the instance belongs is used as an example for explanation. Referring to Figure 8, the method includes:

Step 301: The first routing node receives multiple segment identifiers corresponding to the identifiers of the target computing services advertised by other routing nodes in the computing power network.

In this embodiment of the present application, at least one computing service instance can be deployed in each service node in the computing network, and each service node can be connected to a routing node. Each computing power service instance can provide one computing power service, and the computing power services provided by different computing power service instances can be the same or different.

It can be understood that a routing node in the computing power network can be connected to one or more service nodes, or it can not be connected to a service node and only serve as a message forwarding node. Among them, each routing node connected to a service node can obtain the segment identifier of each computing power service instance deployed in the directly connected service node (that is, each computing power service instance corresponding to the routing node), as well as the segment identifier of each computing power service instance. The identification of the computing power service provided. For example, the routing node may receive the segment identifier of the computing service instance and the identifier of the computing service sent by the service node, or may receive the segment identifier of the computing service instance and the identifier of the computing service issued by the controller.

Optionally, the identifier of each computing power service may be the common prefix of the segment identifier of each corresponding computing power service instance. Alternatively, the identifier of the computing power service and its corresponding segment identifier can also be independent IPv6 addresses.

In this embodiment of the present application, each routing node connected to a service node can notify other routing nodes in the computing power network of the obtained correspondence between the segment identifier of each computing power service instance and the identity of the computing power service. For example, routing nodes may notify each other through a control plane protocol (such as BGP) of the corresponding relationship between the segment identifier of the computing power service instance and the identifier of the computing power service. Based on this, the first routing node in the computing power network can receive multiple segment identifiers corresponding to the identifiers of the target computing power services advertised by other routing nodes. Furthermore, the first routing node may store the corresponding relationship between the identifier of the target computing service and the plurality of segment identifiers.

For example, referring to Figure 9, assume that the segment identifiers of the computing service instances corresponding to routing node 1, routing node 2 and routing node 3 are SID1, SID2 and SID3 respectively, and these three computing service instances are all used to provide the target Computing power service, the identifier of the target computing power service is S1. Correspondingly, the first routing node can receive the segment identifiers SID1, SID2 and SID3 corresponding to the identifier S1 of the target computing service advertised by routing node 1, routing node 2 and routing node 3, and store the identifier S1 and the segment identifier. The corresponding relationship between SID1, SID2 and SID3.

It can be understood that each routing node connected to a service node can notify all other routing nodes in the computing power network of the corresponding relationship between the segment identifier of the computing power service instance and the identifier of the computing power service. Alternatively, each routing node connected to the service node can also advertise the corresponding relationship to only some routing nodes in the computing power network. The part of the routing nodes may be routing nodes connected to the client device.

It can also be understood that the first routing node can also obtain multiple segment identifiers corresponding to the identifier of the target computing service through other methods. For example, the first routing node may receive multiple segment identifiers corresponding to the identifier of the target computing service issued by the controller. Alternatively, the first routing node may be statically configured with multiple segment identifiers corresponding to the identifier of the target computing service.

Step 302: The client device sends an initial request message to the first routing node.

In this embodiment of the present application, if the client device needs to obtain the target computing power service, it can send an initial request message to the first routing node to which it is directly connected. The initial request message includes the identification of the target computing service. For example, the destination address field of the initial request message may include the identification of the target computing service. Moreover, the identifier of the target computing power service may be an IPv6 address.

It can be understood that different computing power services in the computing power network can be distinguished by different identifiers, and each computing power service can be provided by multiple computing power service instances. The identity of the target computing service requested by the client device may be obtained by the client device from the DNS server. For example, the client device can send a domain name request for the target computing power service to the DNS server, and the DNS server can respond to the domain name request and issue the identification of the target computing power service to the client device.

For example, as shown in Figure 9, assuming that the IP address of the client device is C0 and the identifier of the target computing service is S1, the client device can send an initial request message to the first routing node. The source address (SA) in the initial request message is C0, and the destination address (DA) is S1.

It can also be understood that the initial request message may be the first request message (i.e., the first packet) sent by the client device for the target computing power service, and the request message may request and provide computing power for the target computing power service. The service instance creates a session. After the client device creates a session with a computing power service instance that provides the target computing power service, it can obtain the target computing power service through the session. For example, the request message may be a request message for a transmission control protocol (transmission control protocol, TCP) connection, or may be a request message for a quick UDP Internet connection protocol (quick UDP Internet connections, QUIC) connection. Of course, the request message may also be other first packets used to establish a session above the network layer (such as the application layer and transport layer, etc.), which is not limited in the embodiments of the present application.

It can also be understood that before the above step 302, the instance node corresponding to the computing power service instance may publish the route identified by the target computing power service to the first routing node through the routing protocol. The first routing node can then publish the route identified by the target computing power service to the client device and the network where it is located through the routing protocol. Therefore, the client device can send the initial request message to the first routing node based on the learned route. The routing protocol may be BGP or interior gateway protocol (interior gateway protocol, IGP), etc.

Step 303: The first routing node encapsulates the segment list in the initial request message to obtain a computing power request message.

In this embodiment of the present application, after receiving the initial request message sent by the client device, the first routing node can read the identification of the target computing service in the destination address of the initial request message, and determine the identity of the target computing service. Multiple segment identifiers corresponding to the identifier of the force service. Afterwards, the first routing node can encapsulate SRH in the initial request message to obtain the computing power request message. The segment list of the SRH includes multiple segment identifiers corresponding to the identifier of the target computing power service. Based on this, the computing power request message can indicate multiple computing power service instances used to provide the target computing power service through the multiple segment identifiers it carries.

For example, referring to Figure 9, the first routing node may encapsulate the SRH in the initial request message. The segment list of the SRH includes three segment identifiers SID1, SID2 and SID3 corresponding to the identifier S1 of the target computing service.

Optionally, the first routing node may also obtain network performance parameters between it and the instance nodes corresponding to the multiple computing power service instances, and may also encapsulate the obtained network performance parameters into the computing power request message. For example, referring to FIG. 9, the first routing node may encapsulate the network performance parameter into the extended TLV of the SRH. This makes it easy for the instance node that receives the computing power request message to determine whether its corresponding computing power service instance provides the target computing power service to the client device based on the network performance parameters.

For example, the first routing node can detect and obtain network performance parameters between it and each instance node through two-way active measurement protocol (TWAMP) or flow detection. The flow-following detection may refer to in-situ flow information telemetry (IFIT). The network performance parameter may also be called a service level agreement (service level agreement, SLA) message, and the network performance parameter may include at least one of the following parameters: delay, packet loss rate, jitter, etc.

As a first possible example, the order in which the multiple segment identifiers are arranged in the segment list may be randomly determined by the first routing node. That is, after the first routing node determines multiple segment identifiers, it can randomly encapsulate the multiple segment identifiers into the segment list of the computing power request message.

As a second possible example, after obtaining the network performance parameters between the first routing node and each instance node, the first routing node may arrange the segment identifiers in the segment list in order from high to low network performance. That is, if the network performance between the first routing node and the instance node corresponding to the computing service instance indicated by a certain segment identifier is better, the segment identifier will be ranked higher in the segment list. Correspondingly, the instance node corresponding to the segment identifier can receive the computing power request message sooner.

In this second example, since the instance node with better network performance between the first routing node and the first routing node can receive the computing power request message earlier, the instance node with better network performance can be prioritized to detect whether the request message is sent by Its corresponding computing power service instance provides the target computing power service. Since the instance node will also refer to the network performance between it and the first routing node when detecting whether to provide the target computing power service, it can be ensured that each instance node can determine the computing power service instance used to provide the target computing power service as soon as possible. . Furthermore, the efficiency of providing target computing power services to client devices can be effectively improved.

Optionally, for a scenario where the first routing node can obtain the network performance parameters, the first routing node can also encapsulate multiple segment identifiers that meet the encapsulation conditions into the computing power request message based on the network performance parameters. The encapsulation condition may include: the network performance between the first routing node and the instance node corresponding to the computing power service instance indicated by the segment identifier meets the communication condition. That is to say, the first routing node can encapsulate the segment identifier corresponding to the instance node with better network performance into the computing power request message, while the segment identifier corresponding to the instance node with poor network performance does not need to be encapsulated.

Since the first routing node can filter out instance nodes with poor network performance in advance, it can ensure that only instance nodes with better network performance determine whether their corresponding computing power service instance provides the target computing power service. Furthermore, not only the efficiency of determining the computing power service instance that meets the service conditions can be ensured, but also the reliability of providing the target computing power service to the client device can be ensured.

For example, assume that the first routing node obtains multiple candidate segment identifiers corresponding to the identifiers of the target computing power services based on notifications from other routing nodes. Then the first routing node can detect the network performance between it and the instance node corresponding to each candidate segment identifier, and can filter out multiple segment identifiers that meet the encapsulation conditions from the multiple candidate segment identifiers based on the network performance.

The communication condition may include a parameter value range that corresponds one-to-one to each parameter in the network performance parameters. For example, if the network performance parameter includes delay, the communication condition may include a delay range, and the delay range may be a range smaller than a certain delay threshold. If the network performance parameter includes a packet loss rate, the communication condition may include a packet loss rate range, and the packet loss rate range may be a range smaller than a certain packet loss rate threshold. If the network performance parameter includes jitter, the communication condition may include a jitter range, and the jitter range may be a range smaller than a certain jitter threshold.

Step 304: The first routing node forwards the computing power request message according to the segment list.

After the first routing node generates the computing power request message, it can forward the computing power request message according to the segment list. For example, the first routing node may copy the first segment identifier in the segment list to the destination address field of the computing power request message, and forward the computing power request message to the instance node corresponding to the first segment identifier. That is, the first routing node can forward the computing power request message based on the forwarding mechanism of the SRv6 message.

For example, as shown in Figure 9, the first routing node can update the value of the destination address field of the computing power request message to the first segment identifier: SID1, and forward the computing power request message to the instance node corresponding to SID1, That is routing node 1.

Step 305: After determining that the first segment identifier matches the segment identifier of the first computing power service instance, the second routing node detects whether the first computing power service instance satisfies the service condition.

The second routing node may be an instance node corresponding to the computing power service instance indicated by any segment identifier in the segment list. That is, the second routing node may be any instance node that receives the computing power request message in the forwarding path indicated by the segment list. Furthermore, the second routing node corresponds to the first computing power service instance used to provide the target computing power service, that is, the second routing node is directly connected to the service node to which the first computing power service instance belongs.

In this embodiment of the present application, the destination address field in the computing power request message received by the second routing node may include the first segment identifier in the segment list. The second routing node may first detect whether the first segment identifier matches the segment identifier of the first computing power service instance. If the two do not match, the second routing node can continue to forward the computing power request message according to the segment list without checking whether the first computing power service instance meets the service conditions. If the two match, the second routing node can continue to detect whether the first computing power service instance meets the service conditions. If the first computing power service instance does not meet the service conditions, the second routing node can perform step 306; if the first computing power service instance meets the service conditions, the second routing node can perform step 308.

Optionally, the service condition may include: the computing power load of the first computing power service instance is less than the second load threshold. The computing power load can be characterized by at least one of the following parameters: processor usage, memory usage, number of online connections, network delay and integration overhead, etc. The processor usage may include CPU usage and/or GPU usage. The number of online connections can also be called the number of sessions, which refers to the total number of sessions established between the computing service instance and each client device. Network delay refers to the delay between the computing service instance and the first routing node. Fusion overhead refers to a comprehensive metric value calculated based on other parameters in the computing load (such as processor usage, memory usage, number of online connections, and/or network delay), and the comprehensive metric value can be a service node Or calculated by the second routing node. Optionally, the above computing power service instance may refer to the service node corresponding to the computing power service instance.

The second load threshold may include multiple thresholds corresponding to each parameter of the computing power load. For example, the second load threshold may include at least one of the following thresholds: processor usage threshold, memory usage threshold, online connection threshold, network delay threshold, convergence overhead threshold, etc. Correspondingly, the computing power load of the computing power service instance is less than the second load threshold, which may mean: each parameter in the computing power load is less than its corresponding threshold, or the number of parameters in the computing power load that is less than the corresponding threshold is greater than the number. threshold. Based on the above service conditions, it can be ensured that the computing power load of the computing power service instance that provides the target computing power service to the client device is small, thereby ensuring the reliability of providing the target computing power service.

Optionally, the service condition may also include: network performance between the first routing node and the first node (ie, the second routing node) meets the second communication condition. The second communication condition may include a parameter value range corresponding to each parameter in the network performance parameters. For example, the second communication condition may include a delay range, a packet loss rate range, and/or a jitter range. Moreover, the second communication condition may be the same as the communication condition described in step 303, or may be different.

It can be understood that the second routing node can determine the network performance between it and the first routing node based on the network performance parameters encapsulated in the computing power request message. Alternatively, if the network performance parameter is not encapsulated in the computing power request message, the second routing node can also directly detect the network performance between it and the first routing node. For example, the second routing node can be detected through TWAMP or flow detection.

Step 306: The second routing node detects whether the first segment identifier is the last segment identifier in the segment list.

In this embodiment of the present application, if the second routing node detects that the first computing power service instance does not meet the service conditions, it can continue to detect whether the first segment identifier is the last segment identifier in the segment list. If the first segment identifier is not the last segment identifier in the segment list, the second routing node can determine that it can continue to forward the computing power request message to the instance node corresponding to the subsequent segment identifier according to the segment list, and therefore can perform the steps 307. If the first segment identifier is the last segment identifier in the segment list, the second routing node can determine that it is the tail node in the forwarding path indicated by the segment list, and therefore step 308 can be performed.

Step 307: The second routing node determines that the target computing power service is not provided by the first computing power service instance, and forwards the computing power request message according to the segment list.

In the embodiment of this application, if the first computing power service instance does not meet the service conditions and the first segment identifier is not the last segment identifier in the segment list, the second routing node can determine that the first computing power service instance is not used. Provides target computing power services and can continue to forward computing power request messages based on the segment list. For example, the second routing node may update the value of the destination address field of the computing power request message to the second segment identifier in the segment list, and forward the computing power request message to the second node.

Wherein, the second node is an instance node corresponding to the second computing power service instance indicated by the second segment identifier, and the second segment identifier is the next segment identifier adjacent to the first segment identifier in the segment list. That is, the second routing node can forward the computing power request message based on the forwarding mechanism of the SRv6 message. This can facilitate the second node to continue to determine whether the second computing power service instance satisfies the service conditions, and further determine whether the second computing power service instance provides the target computing power service.

For example, as shown in Figure 9, assuming that routing node 1 determines that the first computing power service instance indicated by SID1 does not meet the service conditions, then since the next segment adjacent to SID1 in the segment list is identified as SID2, routing node 1 can Update the value of the destination address field of the computing power request message to SID2 through the SR module. Afterwards, routing node 1 can forward the computing power request message to the second node corresponding to SID2, that is, routing node 2, through the SR module. Among them, the SR module may include software modules and hardware modules that support SRv6 forwarding.

If routing node 2 determines that the second computing power service instance indicated by SID2 also does not meet the service conditions, since the next segment adjacent to SID2 in the segment list is identified as SID3, routing node 2 can report the computing power request through the SR module. The value of the destination address field of the message is updated to SID3. After that, routing node 2 can forward the computing power request message to the instance node corresponding to SID3, that is, routing node 3, through the SR module.

Optionally, before forwarding the computing power request message according to the segment list, the second routing node may also encapsulate the computing power load of the first computing power service instance in the computing power request message. For example, the second routing node may encapsulate the computing power load in the extended TLV of the SRH of the computing power request message, or may encapsulate the computing power load in other extension headers in the computing power request message. Moreover, in order to avoid increasing the data volume of the computing power request message, the second routing node can encapsulate the integration overhead in the computing power load.

By encapsulating the computing power load in the computing power request message, it is convenient for the instance node corresponding to the last segment identifier to determine the computing power load of each computing power service instance, and select the computing power load based on the computing power load of each computing power service instance to provide the client with the computing power load. The terminal device provides the computing power service instance of the target computing power service.

Step 308: The second routing node determines that the first computing power service instance provides the target computing power service, and sends a computing power response message to the client device.

In this embodiment of the present application, if the second routing node detects that the first computing power service instance satisfies the service conditions, it can determine that the first computing power service instance provides the target computing power service, and can send the computing power to the client device. response message. Alternatively, if the second routing node detects that the first computing power service instance does not meet the service conditions and the first segment identifier is the last segment identifier in the segment list, it may also determine that the first computing power service instance provides the target Computing power service, and can send computing power response messages to client devices.

Optionally, the computing power response message may include the first segment identification. For example, the source address field or extended header of the computing power response message includes the first segment identifier. The extension header may be an IPv6 extension header, such as a destination option header (DOH), or a routing extension header (such as SRH). The extended option of the DOH can carry the first segment identifier.

For example, referring to Figure 9, if the second routing node is routing node 3, and routing node 3 determines that the first computing power service instance indicated by SID3 meets the service conditions, it can be determined that the first computing power service instance indicated by SID3 provides Target computing service. Moreover, routing node 3 can send a computing power response message to the client device. The value of the destination address field of the computing power response message may be the IP address C0 of the client device, and the value of the source address field may be the first segment identifier SID3.

It can be understood that the above step 306 can also be executed before step 305, that is, after the second routing node receives the computing power request message and determines that the first segment identifier matches the segment identifier of the first computing power service instance, it can first Checks whether the first segment ID is the last segment ID in the segment list. If the first segment identifier is not the last segment identifier in the segment list, the second routing node can continue to detect whether the first computing power service instance meets the service conditions. If the first segment identifier is the last segment identifier in the segment list, the second routing node can directly perform step 308 without checking whether the first computing power service instance meets the service conditions.

Step 309: The client device sends a computing power task message to the second routing node, and the destination address field of the computing power task message includes the first segment identifier.

After receiving the computing power response message sent by the second routing node, the client device may determine that the target computing power service can be obtained from the first computing power service instance. Therefore, the client device may send the computing power task message to the second routing node, and the destination address field of the computing power task message includes the first segment identification. The second routing node can forward the computing power task message to the service node to which the first computing power service instance belongs, and the first computing power service instance can then provide the target computing power service for the client device.

It can be understood that if the second routing node performs step 307 after the above-mentioned step 305 or step 306, the client device may receive the computing power response message sent by the instance node (such as the routing node) corresponding to the other computing power service instance. arts. Correspondingly, in step 309, the client device may send a computing power task message to instance nodes corresponding to other computing power service instances.

It can also be understood that after receiving the computing power response message, the client device can record the correspondence between the identifier of the target computing power service and the target segment identifier (eg, the first segment identifier) carried in the computing power response message. That is, the client device can learn the target segment identification. Afterwards, if the client device needs to obtain the target computing power service, it can directly send the computing power task message to the target computing power service instance indicated by the target segment identifier based on the corresponding relationship.

Optionally, if the identifier of the target computing power service is the common prefix of multiple segment identifiers corresponding to it, then after receiving the computing power response message, the client device can also first detect whether the prefix of the target segment identifier matches the Matches the identifier of the target computing service. If the two match, the client device can record the correspondence between the identifier of the target computing service and the identifier of the target segment. If the two do not match, the client device does not need to learn the target segment ID. Thus, the target segment identifier can be verified to ensure the reliability of the corresponding relationship recorded by the client device.

It can also be understood that, for the scenario where the first segment identifier is the last segment identifier in the segment list, the above steps 306 and 308 are performed using the second routing node to directly send a computing power response message to the client device as an example. illustrate. As a possible implementation, after the second routing node detects that the first segment identifier is the last segment identifier in the segment list, it may also update the value of the destination address field of the computing power request message to the value in the segment list. The third segment identifies the service instructions. The second routing node may then forward the service indication to the third node.

The third node corresponds to the third computing power service instance indicated by the third segment identifier, and the service indication is used to instruct the third computing power service instance to provide the target computing power service to the client device. That is, if the second routing node detects that the first segment identifier is the last segment identifier in the segment list, it can also forcefully instruct the third computing power service instance to provide the target computing power service through the service indication. The service indication may include an SRH, and the SRH may include a flag bit for mandatory service. For example, the flag bit may be carried in a flag field. Alternatively, the service indication can be obtained by popping the SRH in the computing power request message.

As a first possible example, the third computing power service instance may be randomly selected from multiple computing power service instances indicated by the multiple segment identifiers. For example, the second routing node may randomly select a segment identifier from the segment list as the third segment identifier, and the computing power service instance indicated by the third segment identifier is the third computing power service instance.

As a second possible example, the third computing power service instance may satisfy at least one of the following conditions: the computing power load of the third computing power service instance is less than the first load threshold; the computing power load corresponding to the third computing power service instance Network performance between the third node and the first routing node meets the first communication condition.

In this second example, the computing power request message received by the second routing node may include: the computing power load of the computing power service instance indicated by each segment identifier before the first segment identifier, and/or, in the segment list Network performance parameters between the instance node corresponding to the computing power service instance indicated by each segment identifier and the first routing node. The second routing node may determine the third computing power service instance that satisfies the above conditions based on the computing power load of the multiple computing power service instances and/or the network performance between the multiple instance nodes and the first routing node.

The first communication condition among the above conditions may include a parameter value range corresponding to each parameter in the network performance parameters. For example, the first communication condition may include a delay range, a packet loss rate range, and/or a jitter range, etc.

The first load threshold in the above condition may be a preconfigured load value in the second routing node. Alternatively, the first load threshold may be determined based on the computing power load of the computing power service instance indicated by multiple segment identifiers. For example, the first load threshold may be the mean, median or lower quartile of the computing power loads of the multiple computing power service instances. Alternatively, the first load threshold may be the second smallest value of the computing power load of the multiple computing power service instances. Correspondingly, the service node to which the third computing power service instance belongs is the computing power of the multiple computing power service instances. The service node with the smallest load.

It can be understood that if the second routing node determines multiple alternative computing power service instances whose computing power load is less than the first load threshold, the third computing power service instance can be determined from the multiple alternative computing power service instances. force service instance. For example, the second routing node can randomly select a computing power service instance from multiple alternative computing power service instances as the third computing power service instance, or the second routing node can select a computing power service instance that is in compliance with the network performance of the first routing node. The optimal computing power service instance is used as the third computing power service instance.

The above second example can select a third computing power service instance that meets the conditions based on the computing power load of each computing power service instance and/or the network performance between each instance node and the first routing node. Ensure that the third computing power service instance can provide a relatively reliable target computing power service for the client device.

To sum up, the embodiments of this application provide a computing power request method. After the first node corresponding to the first computing power service instance receives the computing power request message, it can based on whether the first computing power service instance satisfies the service requirements. condition to determine whether the first computing power service instance provides the target computing power service to the client device. Since the first node can directly determine whether its corresponding first computing power service instance meets the service conditions and determine whether to provide the target computing power service, there is no need for routing nodes in the computing power network to centrally sense the computing power of a large number of computing power service instances. load, and select a matching computing service instance. That is to say, the method provided by the embodiments of this application can allow each instance node to determine the computing power service instance used to provide the target computing power service through distributed decision-making, thereby effectively reducing the information release during centralized decision-making by routing nodes. pressure. Furthermore, it effectively reduces the performance requirements for routing nodes and improves the efficiency of providing computing power services.

Figure 10 is a flow chart of yet another computing power request provided by an embodiment of the present application. This method takes the requesting node as a client device and the first node as the first service node to which the first computing power service instance belongs. illustrate. Referring to Figure 10, the method may include:

Step 401: The client device sends a domain name request for the target computing service to the DNS server.

As shown in Figure 11, the domain name request may be the uniform resource locator (URL) of the target computing service.

Step 402: The DNS server sends multiple segment identifiers corresponding to the target computing power service to the client device.

In this embodiment of the present application, after receiving the domain name request sent by the client device, the DNS server can send multiple segment identifiers corresponding to the target computing power service to the client device based on the domain name request. Optionally, the DNS server can also send the identification of the target computing service to the client device.

For example, referring to Figure 11, the DNS server may send three segment identifiers SID1, SID2, and SID3 corresponding to the target computing power service to the client device.

It can be understood that, in addition to obtaining multiple segment identifiers corresponding to the target computing power service through DNS requests, the client device can also obtain the multiple segment identifiers through other methods. For example, the client device may receive multiple segment identifiers corresponding to the target computing power service issued by the controller, or the client device may be statically configured with the multiple segment identifiers in advance.

Step 403: The client device encapsulates the multiple segment identifiers into the computing power request message.

After receiving the multiple segment identifiers corresponding to the target computing power service sent by the DNS server, the client device may record the corresponding relationship between the target computing power service and the multiple segment identifiers. When the client device needs to obtain the target computing power service, it can generate a computing power request message. The computing power request message is encapsulated with SRH, and the segment list of the SRH includes the multiple segment identifiers. For example, as shown in Figure 11, the segment list of the computing power request message contains three segment identifiers, SID1, SID2, and SID3.

Step 404: The client device forwards the computing power request message according to the segment list.

In this embodiment of the present application, the client device can forward the computing power request message according to the segment list based on the forwarding mechanism of the SRv6 message. For example, the client device can copy the first segment identifier in the segment list to the destination address field of the computing power request message, and forward the computing power request message to the instance node corresponding to the first segment identifier.

For example, referring to Figure 11, assuming that the first segment identifier in the segment list is SID1, the value of the destination address field of the computing power request message sent by the client device is SID1. Assume that the segment identifier of the computing power service instance deployed in service node 1 is SID1, then the client device can forward the computing power request message to service node 1 through the routing node.

Step 405: After determining that the first segment identifier matches the segment identifier of the first computing power service instance, the first service node determines whether the first computing power service instance satisfies the service conditions.

Wherein, the first service node may be any service node among the service nodes to which multiple computing power service instances indicated by multiple segment identifiers in the segment list belong, and the first service node is deployed with a device for providing the target computing power service. The first computing power service instance. After the first service node receives the computing power request message, if the destination address field of the computing power request message includes the first segment identifier, it can detect whether the first segment identifier is the same as the segment identifier of the first computing power service instance. match.

If the two do not match, the first service node can continue to forward the computing power request message according to the segment list without checking whether the first computing power service instance meets the service conditions. If the two match, the first service node can continue to detect whether the first computing power service instance meets the service conditions. If the first computing power service instance does not meet the service conditions, the first service node can perform step 406; if the first computing power service instance meets the service conditions, the first service node can perform step 408. For the content of the service conditions, please refer to the relevant description in step 305 above, which will not be described again here.

For example, referring to Figure 11, assume that the first service node is service node 1, and the segment identifier of the first computing power service instance deployed in service node 1 is SID1. If the first segment identifier in the destination address field of the computing power request message received by the service node 1 is also SID1, the service node 1 can detect whether the first computing power service instance meets the service conditions.

Step 406: The first service node detects whether the first segment identifier is the last segment identifier in the segment list.

If the first service node detects that the first computing power service instance does not meet the service conditions, it can continue to detect whether the first segment identifier is the last segment identifier in the segment list. If the first segment identifier is not the last segment identifier in the segment list, the first service node can determine that it can continue to forward the computing power request message to the instance node corresponding to the subsequent segment identifier according to the segment list, so step 407 can be performed. . If the first service node is the last segment identifier in the segment list, the first service node can determine that it is the tail node in the forwarding path indicated by the segment list, and therefore step 408 can be performed.

Step 407: The first service node determines that the target computing power service is not provided by the first computing power service instance, and forwards the computing power request message according to the segment list.

If the first computing power service instance does not meet the service conditions, and the first segment identifier is not the last segment identifier in the segment list, the first service node can determine that the target computing power service is not provided by the first computing power service instance, and can Continue to forward the computing power request message according to this segment list. For example, the first service node may update the value of the destination address field of the computing power request message to the second segment identifier in the segment list, and forward the computing power request message to the second node. The second node may be a service node to which the second computing power service instance indicated by the second segment identifier belongs, and the second segment identifier is the next segment identifier adjacent to the first segment identifier in the segment list.

For example, as shown in Figure 11, assuming that service node 1 determines that the first computing power service instance indicated by SID1 does not meet the service conditions, since the next segment adjacent to SID1 in the segment list is identified as SID2, service node 1 can Update the value of the destination address field of the computing power request message to SID2 through the SR module. Afterwards, service node 1 can forward the computing power request message to the second node corresponding to SID2, that is, service node 2, through the SR module.

If the service node 2 determines that the second computing power service instance indicated by SID2 does not meet the service conditions, since the next segment adjacent to SID2 in the segment list is identified as SID3, the service node 2 can report the computing power request through the SR module. The value of the destination address field of the message is updated to SID3. Afterwards, service node 2 can forward the computing power request message to the instance node corresponding to SID3, that is, service node 3, through the SR module.

Step 408: The first service node determines that the first computing power service instance provides the target computing power service, and sends a computing power response message to the client device.

If the first service node detects that the first computing power service instance meets the service conditions, it can determine that the first computing power service instance provides the target computing power service, and can send a computing power response message to the client device. Alternatively, if the first service node detects that the first computing power service instance does not meet the service conditions and the first segment identifier is the last segment identifier in the segment list, it may also determine that the first computing power service instance provides the target computing power service and sends a computing power response message to the client device. For the implementation process of step 408, please refer to the relevant description of step 308 above.

For example, referring to Figure 11, if the first service node is service node 3, and the value of the destination address field in the computing power request message received by service node 3 is SID3. If the service node 3 determines that the first computing power service instance indicated by SID3 satisfies the service conditions, it may determine that the first computing power service instance indicated by SID3 provides the target computing power service. Moreover, the service node 3 can send a computing power response message to the client device. The value of the destination address field of the computing power response message may be the IP address C0 of the client device, and the value of the source address field may be the first segment identifier SID3.

Step 409: The client device sends a computing power task message to the first service node, and the destination address field of the computing power task message includes the first segment identifier.

After receiving the computing power response message sent by the first service node, the client device may determine that the target computing power service can be obtained from the first computing power service instance deployed in the first service node. Therefore, the client device can record the corresponding relationship between the identifier of the target computing power service and the target segment identifier (eg, the first segment identifier) carried in the computing power response message. Afterwards, the client device can send a computing power task message to the first service node, and the destination address field of the computing power task message includes the first segment identifier. For the implementation process of step 409, please refer to the relevant description of step 309 above.

For the scenario where the instance node (for example, the first node) corresponding to the computing power service instance is a service node, referring to Figure 11, it can be seen that the service node includes both the computing power service module and the SR module that supports SRv6 technology. That is to say, the service node can not only provide computing power services to client devices through computing power service instances, but also forward computing power request messages through the SRv6 message forwarding mechanism.

Comparing the embodiments shown in Figure 8 and Figure 10, it can be seen that if the instance node (for example, the first node) corresponding to the computing power service instance is a service node, the service node can be based on the computing power load of the first computing power service instance deployed by itself. Determine whether the first computing power service instance meets the service conditions. As a result, routing nodes in the computing power network no longer need to sense the computing power load of the computing power service instance and only need to forward messages, thus effectively reducing the control surface pressure of routing nodes.

If the instance node corresponding to the computing power service instance is a routing node (such as a second routing node) directly connected to the service node, the routing node can only sense the computing power of the computing power service instance deployed in the service node to which it is directly connected. load, and the service node does not need to parse the SRH and forward the message according to the segment list. As a result, the control surface pressure of the routing node can be effectively reduced while being effectively compatible with the functions of the traditional service node, that is, the performance requirements for the service node will not be increased.

Comparing the embodiments shown in Figure 8 and Figure 10, it can also be seen that if the requesting node is a routing node directly connected to the client device (for example, the first routing node), the client device does not need to sense and encapsulate the identity of the target computing service. The corresponding multiple segment identifiers do not need to support SRv6 technology. As a result, the performance requirements for client devices are effectively reduced.

If the requesting node is a client device, the routing node directly connected to the client device does not need to sense and encapsulate multiple segment identifiers corresponding to the identifier of the target computing service, but only needs to forward the message based on the forwarding mechanism of the SRv6 message. Can. As a result, the performance requirements for routing nodes are effectively reduced.

The above steps 309 and 409 are explained by taking the client device learning the target segment identifier in the computing power response message as an example. It can be understood that if the client device does not have the learning capability, a routing node (such as the first routing node) between the client device and the service node can also learn the target segment in the computing power response message. logo. That is, the routing node can record the correspondence between the identification of the target computing power service and the identification of the target segment. This correspondence can also be called a flow table. After receiving the computing power request message or computing power task message sent by the client device, the routing node can search the flow table according to the identification of the target computing power service in the message, and can forward it according to the found target segment identification. message.

Optionally, in this embodiment of the present application, the identity of the target computing power service may also be encapsulated in the computing power request message generated by the client device or the first routing node. For example, referring to Figures 9 and 11, the last segment identification field in the segment list of the computing power request message, that is, segment list[0], can carry the identification of the target computing power service (such as S1). This makes it easier for the routing node or service node in the computing power network to determine the source address and destination of the computing power request message when the destination address field in the computing power request message is updated to identify the target computing power service. The address can pass the verification during verification to avoid discarding the computing power request message due to verification failure. The verification method used by the routing node or service node may be TCP or UDP verification and checksum.

It can be understood that in the embodiment shown in FIG. 8 , the client device can also generate a computing power request message based on the method shown in steps 401 to 403 . In the embodiment shown in FIG. 10 , the first routing node may also generate a computing power request message based on the method shown in steps 301 to 303 . That is to say, the requesting node and the first node can be flexibly combined according to the requirements of the application scenario, which is not limited in the embodiments of the present application.

It can also be understood that the sequence of steps in the computing power request method provided by the embodiments of the present application can be appropriately adjusted, and the steps can also be increased or decreased accordingly according to the situation. For example, the above step 301 can be deleted according to the situation. Alternatively, the above step 306 may be deleted according to circumstances, or may be performed before step 305. Or, the above steps 401 and 402 can be deleted according to the situation. Or, the above step 406 can be deleted according to the situation, or can be performed before step 405.

To sum up, the embodiments of this application provide a computing power request method. After the first node corresponding to the first computing power service instance receives the computing power request message, it can based on whether the first computing power service instance satisfies the service requirement. condition to determine whether the first computing power service instance provides the target computing power service to the client device. Since the first node can directly determine whether its corresponding first computing power service instance meets the service conditions and determine whether to provide the target computing power service, there is no need for routing nodes in the computing power network to centrally sense the computing power of a large number of computing power service instances. load, and select a matching computing service instance. That is to say, the method provided by the embodiments of this application can allow each instance node to determine the computing power service instance used to provide the target computing power service through distributed decision-making, thereby effectively reducing the information release during centralized decision-making by routing nodes. pressure. Furthermore, it effectively reduces the performance requirements for routing nodes and improves the efficiency of providing computing power services.

Figure 12 is a flow chart of yet another computing power request method provided by an embodiment of the present application. As shown in Figure 12, the method includes:

Step S1: The instance node receives the computing power request message from the client device.

The segment list of the computing power request message includes multiple segment identifiers, and the multiple segment identifiers respectively indicate multiple computing power service instances, where each computing power service instance is used to provide the target computing power service. The instance node corresponds to one computing power service instance among the multiple computing power service instances, and the computing power service instance corresponding to the instance node can be the computing power indicated by any segment identifier in the segment list except the last segment identifier. Service instance. Optionally, the instance node may be a service node to which its corresponding computing power service instance belongs, or may be a routing node directly connected to the service node.

Step S2: The instance node encapsulates the computing power load of its corresponding computing power service instance in the computing power request message.

In this embodiment of the present application, after receiving the computing power request message, the instance node can encapsulate the computing power load of its corresponding computing power service instance into the computing power request message. For example, the instance node can encapsulate the computing power load (such as the integration overhead) into the extended TLV of the SRH of the computing power request message.

Step S3: The instance node forwards the computing power request message according to the segment list.

The instance node can forward the computing power request message according to the segment list according to the SRv6 message forwarding mechanism. It can be understood that the instance node corresponding to the computing service instance indicated by each segment identifier except the last segment identifier in the segment list can execute the method shown in steps S1 to S3 above.

Step S4: The tail instance node sends a service instruction to the instance node corresponding to the target computing service instance based on the computing power load of the multiple computing power service instances.

The tail instance node is the instance node corresponding to the computing service instance indicated by the last segment identifier in the segment list. Based on the above steps S1 to S3, it can be known that the computing power request message received by the tail instance node also includes: other computing power service instances among the multiple computing power service instances except the computing power service instance corresponding to the tail instance node. computing power load. Therefore, the tail instance node can send a service instruction to the instance node corresponding to the target computing power service instance based on the computing power load of the multiple computing power service instances, so that the target computing power service instance provides the target computing power to the client device. Serve.

Wherein, the target computing power service instance can satisfy: the computing power load of the target computing power service instance is less than the first load threshold.

Optionally, the target computing power service instance may also satisfy: the network performance between the instance node corresponding to the target computing power service instance and the first routing node meets the first communication condition. Wherein, the first routing node is a routing node connected to the client device.

It can be understood that the computing power request message may also include: network performance parameters between the first routing node and the instance nodes corresponding to the multiple computing power service instances, and the tail instance node may determine based on the network performance parameters that satisfy The target computing power service instance of the first communication condition.

The network performance parameter may be detected by the first routing node and encapsulated into the computing power request message. Alternatively, among the multiple computing power service instances, the instance nodes corresponding to other computing power service instances except the computing power service instance corresponding to the tail instance node can detect and encapsulate the network performance parameters between them and the first routing node. .

It can also be understood that the above-mentioned step S2 can be deleted according to the situation. Correspondingly, in the above-mentioned step S4, the tail instance node can determine the target computing power service instance from multiple computing power service instances based on other methods. For example, the tail instance node can randomly select a computing power service instance from multiple computing power service instances as the target computing power service instance. Alternatively, the tail instance node can determine the target computing service instance based on network performance parameters.

For explanations about the above-mentioned first load threshold and first communication condition, reference may be made to the relevant description after step 309 in the embodiment shown in FIG. 8 , which will not be described again here.

Figure 13 is a schematic structural diagram of a first node provided by an embodiment of the present application. The first node can be applied to a computing power network such as that shown in Figure 1, and the first node corresponds to a first computing power service instance, and the first computing power service instance is used to provide a target computing power service. For example, the first node may be a service node where the first computing power service instance is deployed, or may be a routing node directly connected to the service node. Moreover, the first node can implement the steps performed by the first node or the second routing node in the above method embodiment. Referring to Figure 13, the first node includes:

The receiving module 501 is configured to receive a computing power request message from a client device, where the computing power request message is used to request a target computing power service. For the functional implementation of the receiving module 501, please refer to the relevant descriptions of step 101, step 304 and step 404 in the above method embodiment.

The determination module 502 is configured to determine whether the first computing power service instance provides the target computing power service to the client device based on whether the first computing power service instance satisfies the service condition. For the functional implementation of the determination module 502, please refer to the relevant descriptions of step 102, step 305 and step 405 in the above method embodiment.

Optionally, the segment list of the computing power request message includes multiple segment identifiers; the multiple segment identifiers respectively indicate multiple computing power service instances, and the multiple computing power service instances are respectively used to provide target computing power services. , the plurality of computing power service instances include the first computing power service instance.

Optionally, as shown in Figure 13, the first node may also include a sending module 503.

The determination module 502 may be used to determine that the first computing power service instance provides the target computing power service to the client device based on the first computing power service instance satisfying the service condition, and send the target computing power service to the client device through the sending module 503. Send a computing power response message; or, based on the fact that the first computing power service instance does not meet the service conditions, determine that the first computing power service instance does not provide the target computing power service to the client device, and use the sending module 503 according to the segment list Forward the computing power request message.

The functional implementation of the determining module 502 and the sending module 503 may also refer to the relevant descriptions of steps 307, 308, 407 and 408 in the above method embodiment.

Optionally, the destination address field of the computing power request message includes a first segment identifier used to indicate the first computing power service instance, and the plurality of segment identifiers include the first segment identifier. The sending module 503 may be configured to: update the value of the destination address field of the computing power request message to the second segment identifier in the segment list, and forward the computing power request message to the second node. The second node corresponds to the second computing power service instance indicated by the second segment identifier, and the second segment identifier is the next segment identifier adjacent to the first segment identifier in the segment list.

Optionally, the sending module 503 may also be configured to encapsulate the computing power load of the first computing power service instance in the computing power request message.

Optionally, the destination address field of the computing power request message includes a first segment identifier used to indicate the first computing power service instance, and the plurality of segment identifiers include the first segment identifier. The sending module 503 can be used to: based on the first segment identifier being the last segment identifier in the segment list, update the value of the destination address field of the computing power request message to the third segment identifier in the segment list to obtain service Instruction; forward the service instruction to a third node, the third node corresponds to the third computing power service instance indicated by the third segment identifier, and the service instruction is used to instruct the third computing power service instance to provide the target to the client device Computing service.

Optionally, the third computing power service instance is randomly selected from multiple computing power service instances; or, the third computing power service instance meets at least one of the following conditions: The computing power load is less than the first load threshold; the network performance between the third node corresponding to the third computing power service instance and the first routing node meets the first communication condition, and the first routing node is a routing node connected to the client device .

Optionally, the determining module 502 may be configured to determine, based on the first segment identifier being the last segment identifier in the segment list, to provide the target computing power service to the client device by the first computing power service instance, and send the Module 503 sends a computing power response message to the client device. For the functional implementation of the determination module 502, reference may also be made to the relevant descriptions of steps 306 and 406 in the above method embodiments.

Optionally, the computing power response message may include the first segment identification. For example, the source address field or extended header of the computing power response message includes the first segment identifier.

Optionally, the segment list may also include the identification of the target computing power service.

Optionally, the service condition may include: the computing power load of the first computing power service instance is less than the second load threshold. The service condition may further include: network performance between the first routing node connected to the client device and the first node meets the second communication condition.

To sum up, the embodiments of the present application provide a first node. After receiving the computing power request message, the first node can determine whether to use the computing power request message based on whether the corresponding first computing power service instance satisfies the service conditions. The first computing power service instance provides the target computing power service to the client device. Since the first node can directly determine whether its corresponding first computing power service instance meets the service conditions and determine whether to provide the target computing power service, there is no need for routing nodes in the computing power network to centrally sense the computing power of a large number of computing power service instances. load, and select a matching computing service instance. That is to say, the first node and the nodes corresponding to other computing power service instances can determine the computing power service instance used to provide the target computing power service through distributed decision-making, thereby effectively reducing the information in the centralized decision-making process by the routing node. Pressure to publish. Furthermore, it effectively reduces the performance requirements for routing nodes and improves the efficiency of providing computing power services.

Figure 14 is a schematic structural diagram of a request node provided by an embodiment of the present application. The request node may be applied to a computing power network such as that shown in Figure 1, and the request node may be a client device, or may be a routing node directly connected to the client device. Moreover, the requesting node may implement the steps performed by the requesting node, the client device or the first routing node in the above method embodiment. Referring to Figure 14, the request node includes:

The generation module 601 is configured to generate a computing power request message, where the segment list of the computing power request message includes multiple segment identifiers. The multiple segment identifiers respectively indicate multiple computing power service instances, and each of the multiple computing power service instances is used to provide a target computing power service. For the functional implementation of the generation module 601, please refer to the relevant descriptions of step 201 and step 403 in the above method embodiment.

The sending module 602 is configured to forward the computing power request message according to the segment list, so that at least one computing power service instance among the plurality of computing power service instances that meets the service condition provides the target computing power service to the client device. For the functional implementation of the sending module 602, please refer to the relevant descriptions of steps 202 and 404 in the above method embodiment.

Optionally, for a scenario where the requesting node is a client device, the generation module 601 may be configured to encapsulate multiple segment identifiers corresponding to the target computing service based on the target computing service requested by the client device. to the computing power request message.

Optionally, the sending module 602 can also be used to send a domain name request for the target computing power service to the DNS server. Moreover, as shown in Figure 14, the client device may also include a receiving module 603, configured to receive multiple segment identifiers corresponding to the target computing power service sent by the DNS server.

The functional implementation of the sending module 602 may also refer to the relevant description of step 401 in the above method embodiment. For the functional implementation of the receiving module 603, please refer to the relevant description of step 402 in the above method embodiment.

Optionally, for the scenario where the requesting node is a routing node connected to the client device, the generating module 601 may be used to receive an initial request message sent by the client device, where the initial request message includes the target computing service. The identifier; encapsulate multiple segment identifiers corresponding to the identifier of the target computing power service in the initial request message to obtain the computing power request message. For the functional implementation of the generation module 601, you may also refer to the relevant description of step 303 in the above method embodiment.

Optionally, the receiving module 603 may be configured to receive multiple segment identifiers corresponding to the identifier of the target computing service advertised by other routing nodes.

Optionally, the generation module 601 may also be configured to encapsulate network performance parameters between the routing node and the instance node corresponding to the computing power service instance indicated by the multiple segment identifiers into the computing power request message.

Optionally, the computing power request message includes an SRH, and the extended TLV field in the SRH includes the network performance parameter.

Optionally, the generation module 601 may also be used to filter out multiple segment identifiers that meet the encapsulation conditions from multiple candidate segment identifiers corresponding to the identifier of the target computing power service. The encapsulation conditions include: the network performance between the routing node and the instance node corresponding to the computing power service instance indicated by the segment identifier meets the communication conditions.

Optionally, the multiple segment identifiers are arranged in order from high to low network performance.

Optionally, the receiving module 603 can also be used to receive a computing power response message sent by the target node. The computing power response message includes the target segment identifier of the target computing power service instance corresponding to the target node. The target segment The identifier belongs to the multiple segment identifiers. For the functional implementation of the receiving module 603, reference may also be made to the relevant descriptions of steps 308 and 408 in the above method embodiment.

The sending module 602 may also be used to send a computing power task message to the target node, where the destination address field of the computing power task message includes the target segment identifier. For the functional implementation of the sending module 602, reference may also be made to the relevant descriptions of steps 308 and 408 in the above method embodiment.

Optionally, the source address field or extension header of the computing power response message includes the target segment identifier.

Optionally, the identifier of the target computing power service is a common prefix of the multiple segment identifiers. The sending module 602 may also be configured to record the corresponding relationship between the target computing service and the target segment identifier based on the match between the prefix of the target segment identifier and the identifier of the target computing service.

Optionally, the segment list may also include the identification of the target computing power service.

To sum up, embodiments of the present application provide a requesting node. The computing power request message generated by the requesting node includes a segment list, and the computing power request message can be forwarded according to the segment list. Since each segment identifier in the segment list indicates a computing power service instance used to provide the target computing power service, the computing power request message can be forwarded to the corresponding computing power service instance along the forwarding path indicated by the segment list. Instance node. Furthermore, the instance node that receives the computing power request message can determine whether the computing power service instance provides the target computing power service to the client device based on whether its corresponding computing power service instance meets the service conditions. As a result, routing nodes in the computing network no longer need to centrally sense the computing load of a large number of computing service instances and select matching computing service instances, thereby effectively reducing the performance requirements for routing nodes and improving the provision of computing power. Efficiency in service.

Those skilled in the art can clearly understand that for the convenience and simplicity of description, the specific working processes of the first node, the request node and each module described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here. Repeat.

It should be understood that the first node and the requesting node provided by the embodiment of the present application can also be implemented by an application-specific integrated circuit (ASIC) or a programmable logic device (PLD), as mentioned above. PLD can be a complex programmable logical device (CPLD), a field-programmable gate array (FPGA), a general array logic (GAL), or any combination thereof. In addition, the computing power requesting method provided by the above method embodiment can also be implemented through software. When the computing power requesting method provided by the above method embodiment is implemented through software, each functional module in the first node and the requesting node can also be software. module.

Figure 15 is a schematic structural diagram of a computing power request device provided by an embodiment of the present application. The computing power requesting device can be applied to a computing power network such as that shown in Figure 1, and can be the first node or requesting node in the network. That is, the computing power requesting device may be a client device, a routing node or a service node. Referring to Figure 15, the computing power requesting device includes: a processor 701, a memory 702, a network interface 703 and a bus 704.

Among them, the computer program 7021 is stored in the memory 702, and the computer program 7021 is used to implement various application functions. The processor 701 is configured to execute the computer program 7021 to implement the method provided by the above method embodiment and applied to the first node, the second node or the third node. For example, the processor 701 can be used to execute the computer program 7021 to implement the functions of each module in the node shown in Figure 13 or Figure 14.

The processor 701 may be a central processing unit (CPU). The processor 701 may also be other general-purpose processors, digital signal processors (DSP), ASICs, FPGAs, or graphics processing units. unit, GPU) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general purpose processor can be a microprocessor or any conventional processor.

Memory 702 may be volatile memory or non-volatile memory, or may include both volatile and non-volatile memory. Among them, the non-volatile memory can be read-only memory (ROM), programmable ROM (PROM), erasable programmable read-only memory (erasable PROM, EPROM), electrically removable memory. Erase programmable read-only memory (electrically EPROM, EEPROM) or flash memory. Volatile memory may be random access memory (RAM), which is used as an external cache. By way of illustration, but not limitation, many forms of RAM are available, such as static random access memory (SRAM), dynamic random access memory (DRAM), synchronous dynamic random access memory (synchronous DRAM, SDRAM), double data rate synchronous dynamic random access memory (double data date SDRAM, DDR SDRAM), enhanced synchronous dynamic random access memory (enhanced SDRAM, ESDRAM), synchronous link dynamic random access memory (synchlink DRAM, SLDRAM) ) and direct memory bus random access memory (direct rambus RAM, DRRAM).

There may be multiple network interfaces 703 , and the network interfaces 703 are used to implement communication connections with other devices (which may be wired or wireless). Among them, in this embodiment of the present application, the network interface 703 is used to send and receive messages. Among them, other devices may be terminals, servers, VMs and other devices or other network devices.

The bus 704 is used to connect the processor 701, the memory 702 and the network interface 703. Furthermore, in addition to the data bus, the bus 704 may also include a power bus, a control bus, a status signal bus, etc. However, for the sake of clarity, the various buses are labeled bus 704 in the figure.

If the computing power requesting device is a first node, the first node corresponds to a first computing power service instance, and the first computing power service instance is used to provide the target computing power service. The processor 701 is configured to receive a computing power request message from the client device through the network interface 703. The computing power request message is used to request a target computing power service, and is used to determine whether the first computing power service instance satisfies the service requirement. condition to determine whether the first computing power service instance provides the target computing power service to the client device. For the detailed processing process of the processor 701, please refer to the steps performed by the first node (for example, the second routing node or the first service node) in the method embodiments shown in FIG. 6, FIG. 8, and FIG. 10, which will not be described again here.

If the computing power requesting device is a requesting node, the processor 701 is configured to generate a computing power request message. The segment list of the computing power request message includes multiple segment identifiers, and the multiple segment identifiers respectively indicate multiple computing power requests. A computing power service instance, each of the multiple computing power service instances is used to provide a target computing power service. The processor 701 is also configured to forward the computing power request message through the network interface 703 according to the segment list, so that at least one computing power service instance among the multiple computing power service instances that meets the service conditions is forwarded to the client device. Provide target computing power services. For the detailed processing process of the processor 701, please refer to the steps performed by the requesting node (such as the client device or the first routing node) in the method embodiments shown in FIG. 7, FIG. 8 and FIG. 10, which will not be described again here.

Figure 16 is a schematic structural diagram of another computing power requesting device provided by an embodiment of the present application. The computing power requesting device can be applied to a computing power network such as that shown in Figure 1, and can be the first node or requesting node in the network. That is, the computing power requesting device may be a client device, a routing node or a service node. As shown in Figure 16, the computing power requesting device may include: a main control board 801 and at least one interface board (the interface board is also called a line card or a service board). For example, Figure 16 shows the interface board 802 and the interface board 803. . In the case of multiple interface boards, the computing power requesting device may also include a switching network board 804, which is used to complete data exchange between the interface boards.

The main control board 801 is also called a main processing unit (MPU) or a route processor card. The main control board 801 is used to complete functions such as system management, device maintenance, and protocol processing. There are three main types of functional units on the main control board 801: system management control unit, system clock unit and system maintenance unit. The main control board 801 includes: a central processing unit 8011 and a memory 8012.

The interface boards 802 and 803 are also called line processing unit (LPU), line card (line card) or service boards. The interface boards are used to provide various service interfaces and implement packet forwarding. Among them, the service interfaces provided by the interface board may include: SONET/SDH-based packet (packet over SONET/SDH, POS) interface, Gigabit Ethernet (gigabit Ethernet, GE) interface and asynchronous transfer mode (asynchronous transfer mode), ATM) interface, etc. Among them, SONET refers to synchronous optical network (synchronous optical network), and SDH refers to synchronous digital hierarchy (synchronous digital hierarchy). The main control board 801, the interface board 802 and the interface board 803 are connected to the system backplane through a system bus to achieve intercommunication. As shown in Figure 16, the interface board 802 includes one or more central processors 8021. The central processor 8021 is used to control and manage the interface board 802 and communicate with the central processor 8011 on the main control board 801 . The memory 8024 on the interface board 802 is used to store forwarding entries, and the network processor 8022 can forward messages by searching for the forwarding entries stored in the memory 8024. Memory 8024 may also be used to store program code.

The interface board 802 also includes one or more physical interface cards 8023. The one or more physical interface cards 8023 are used to receive messages sent by the previous hop node and send them to the next hop node according to the instructions of the central processor 8021. processed message.

In addition, it can be understood that the central processor 8021 and/or the network processor 8022 in the interface board 802 in Figure 16 can be dedicated hardware or chips. For example, an ASIC can be used to implement the above functions. This implementation method is generally The forwarding plane is processed by dedicated hardware or chips. In other embodiments, the central processor 8021 and/or the network processor 8022 may also use a general-purpose processor, such as a general-purpose CPU, to implement the functions described above.

In addition, it should be understood that there may be one or more main control boards 801, and when there are multiple main control boards, they may include a main main control board and a backup main control board. There may be one or more interface boards. The stronger the data processing capability of the device required by the computing power, the more interface boards are provided. As shown in Figure 16, the computing power requesting device includes an interface board 802 and an interface board 803. When a distributed forwarding mechanism is adopted, the structure of the interface board 803 is basically the same as that of the interface board 802, and the operations on the interface board 803 are basically similar to the operations on the interface board 802. For the sake of simplicity, they will not be described again. When the computing power requesting device has multiple interface boards, the multiple interface boards can communicate with each other through one or more switching network boards 804, and can realize load sharing and redundant backup to provide large-capacity data exchange and processing power.

Under the centralized forwarding architecture, the computing power requesting device does not need the switching network board 804, and the interface board is responsible for processing the business data of the entire system. Therefore, the data access and processing capabilities of computing power requesting equipment in a distributed architecture are greater than those in a centralized architecture. The specific architecture used depends on the specific networking deployment scenario and is not limited here.

In this embodiment of the present application, the memory 8012 and the memory 8024 can be ROM or other types of static storage devices that can store static information and instructions, or RAM or other types of dynamic storage devices that can store information and instructions. They can also be It is EEPROM, compact disc read-only memory (CD-ROM) or other optical disc storage, optical disc storage (including compressed optical disc, laser disc, optical disc, digital versatile disc, Blu-ray disc, etc.), magnetic disk or other magnetic storage device , or any other medium that can be used to carry or store desired program code in the form of instructions or data structures and that can be accessed by a computer, but is not limited thereto. The memory 8024 in the interface board 802 can exist independently and be connected to the central processor 8021 through a communication bus; or the memory 8024 can also be integrated with the central processor 8021. The memory 8012 in the main control board 801 can exist independently and be connected to the central processor 8011 through a communication bus; or the memory 8012 can also be integrated with the central processor 8011.

The program code stored in the memory 8024 is controlled and executed by the central processor 8021, and the program code stored in the memory 8012 is controlled and executed by the central processor 8011. The central processor 8021 and/or the central processor 8011 can implement the method executed by any node provided by the above method embodiments by executing program codes. One or more software units may be included in the program code stored in memory 8024 and/or memory 8012. This one or more software units may be functional modules shown in any of the figures in Figures 13 to 15.

In this embodiment of the present application, the physical interface card 8023 can be a device using any transceiver, used to communicate with other devices or communication networks, such as Ethernet, radio access network (radio access network, RAN), Wireless local area networks (WLAN), etc.

Optionally, the device shown in any one of Figures 13 to 15 can also be implemented using the structure shown in Figure 16 .

Embodiments of the present application also provide a computer-readable storage medium. The computer-readable storage medium stores instructions. When the instructions are run on the processor, the processor executes the steps performed by the first node or the processor in the above method embodiment. The steps to be performed by the requesting node.

Embodiments of the present application also provide a computer program product containing instructions. When the instructions are run on a processor, they cause the processor to perform the steps performed by the first node or the requesting node in the above method embodiments.

Embodiments of the present application also provide a computing power request system. The computing power request system includes multiple instance nodes. At least one instance node among the multiple instance nodes can be used to implement the method executed by the first node in the above method embodiment. Methods. For example, referring to Figure 1, the routing node 01 or the service node 02 in the computing power request system can be used to implement the method performed by the first node in the above method embodiment.

Optionally, each instance node among the plurality of instance nodes may be a service node to which its corresponding computing power service instance belongs; or, each instance node among the plurality of nodes may be a service node to which its corresponding computing power service instance belongs. The routing node directly connected to the service node to which the instance belongs.

Optionally, the system may also include: a request node, which may be used to implement the method performed by the request node in the above method embodiment.

Optionally, referring to FIG. 1 , the requesting node may be the client device 03 , or the requesting node may be a routing node connected to the client device 03 .

The structure of the instance node may be as shown in Figure 13, Figure 15 or Figure 16, and the structure of the request node may be as shown in Figure 14, Figure 15 or Figure 16.

Those of ordinary skill in the art can understand that all or part of the steps to implement the above embodiments can be completed by hardware, or can be completed by instructing relevant hardware through a program. The program can be stored in a computer-readable storage medium. The above-mentioned The storage media mentioned can be read-only memory, magnetic disks or optical disks, etc.

In the embodiments of the present application, the terms "first", "second" and "third" are only used for description purposes and cannot be understood as indicating or implying relative importance. The term "and/or" in this application is just an association relationship describing related objects, indicating that there can be three relationships, for example, A and/or B, which can mean: A exists alone, A and B exist simultaneously, alone There are three situations B. In addition, the character "/" in this article generally indicates that the related objects are an "or" relationship.

The above are only optional embodiments of the present application and are not intended to limit the present application. Any modifications, equivalent substitutions, improvements, etc. made within the concepts and principles of the present application shall be included in the protection of the present application. within the range.

Claims (39)

1. A method of computing a force request, applied to a first node, the method comprising:

receiving an account force request message from a client device, wherein the account force request message is used for requesting a target account force service, the first node corresponds to a first account force service instance, and the first account force service instance is used for providing the target account force service;

based on whether the first computing force service instance satisfies a service condition, it is determined whether the target computing force service is provided by the first computing force service instance to the client device.

2. The method of claim 1, wherein the segment list of the power request message includes a plurality of segment identifiers;

the plurality of segment identifications respectively indicate a plurality of computing force service instances, and the plurality of computing force service instances respectively are used for providing the target computing force service, and the plurality of computing force service instances comprise the first computing force service instance.

3. The method of claim 2, wherein the determining whether to provide the target computing force service to the client device by the first computing force service instance based on whether the first computing force service instance satisfies a service condition comprises:

Determining that the first computing power service instance provides the target computing power service to the client device based on the first computing power service instance meeting the service condition, and sending a computing power response message to the client device; or,

and determining that the target computing power service is not provided to the client device by the first computing power service instance based on that the first computing power service instance does not meet the service condition, and forwarding the computing power request message according to the segment list.

4. A method according to claim 3, wherein the destination address field of the power request message comprises a first segment identification indicating the first power service instance, the plurality of segment identifications comprising the first segment identification; the forwarding the power calculation request message according to the segment list comprises the following steps:

updating the value of the destination address field of the calculation request message to be a second segment identifier in the segment list;

forwarding the calculation force request message to a second node, wherein the second node corresponds to a second calculation force service instance indicated by the second segment identification;

wherein the second segment identifier is a next segment identifier adjacent to the first segment identifier in the segment list.

5. The method according to claim 3 or 4, wherein before forwarding the power request message according to the segment list, the method further comprises:

and encapsulating the computational load of the first computational service instance in the computational request message.

6. A method according to claim 3, wherein the destination address field of the power request message comprises a first segment identification indicating the first power service instance, the plurality of segment identifications comprising the first segment identification; the forwarding the power calculation request message according to the segment list comprises the following steps:

updating the value of the destination address field of the calculation request message to be a third segment identifier in the segment list based on the first segment identifier as the last segment identifier in the segment list to obtain a service instruction;

forwarding the service indication to a third node, the third node corresponding to a third computing force service instance of the third segment identification indication, and the service indication for indicating the third computing force service instance to provide the target computing force service to the client device.

7. The method of claim 6, wherein the third computing force service instance is randomly selected from the plurality of computing force service instances; alternatively, the third computing force service instance satisfies at least one of the following conditions:

The computational load of the third computational service instance is less than a first load threshold;

the network performance between a third node corresponding to the third computing power service instance and a first routing node meets a first communication condition, and the first routing node is a routing node connected with the client device.

8. The method of claim 2, wherein the determining whether to provide the target computing force service to the client device by the first computing force service instance based on whether the first computing force service instance satisfies a service condition comprises:

determining that the first computing power service instance provides the target computing power service to the client device based on the first segment identifier as the last segment identifier in the segment list, and sending a computing power response message to the client device.

9. The method according to any one of claims 3 to 8, wherein the first segment identification is included in the force response message.

10. The method of claim 9, wherein the first segment identifier is included in a source address field or an extension header of the power response message.

11. The method according to any of claims 3 to 10, wherein the segment list further comprises an identification of the target computing force service.

12. The method according to any one of claims 1 to 11, wherein the service conditions comprise: the computational load of the first computational service instance is less than a second load threshold.

13. The method of claim 12, wherein the service conditions further comprise: network performance between a first routing node to which the client device is connected and the first node satisfies a second communication condition.

14. The method according to any one of claims 1 to 13, wherein the first node is a service node to which the first computing power service instance belongs;

or the first node is a second routing node directly connected with the service node to which the first computing power service instance belongs.

15. A method of computing a force request, the method comprising:

generating a power calculation request message, wherein a segment list of the power calculation request message comprises a plurality of segment identifiers, the segment identifiers respectively indicate a plurality of power calculation service instances, and each power calculation service instance in the plurality of power calculation service instances is used for providing a target power calculation service;

and forwarding the computing power request message according to the segment list, so that at least one computing power service instance meeting service conditions in the plurality of computing power service instances provides the target computing power service for the client device.

16. The method of claim 15, wherein the computing force request method is applied to the client device; the generating the calculation force request message comprises the following steps:

and based on the target computing power service requested by the client device, packaging the section identifiers corresponding to the target computing power service into the computing power request message.

17. The method of claim 16, wherein prior to the generating the power request message, the method further comprises:

sending a domain name request for the target computing service to a domain name system DNS server;

and receiving the section identifiers corresponding to the target computing power service, which are sent by the DNS server.

18. The method of claim 15, wherein the power request method is applied to a routing node connected to the client device; the generating the calculation force request message comprises the following steps:

receiving an initial request message sent by the client device, wherein the initial request message comprises an identifier of the target computing power service;

and encapsulating the section identifiers corresponding to the identifiers of the target computing service in the initial request message to obtain the computing request message.

19. The method of claim 18, wherein prior to the generating the power request message, the method further comprises:

and receiving the segment identifications corresponding to the identifications of the target computing power service advertised by other routing nodes.

20. The method of claim 18 or 19, wherein generating the computational power request message further comprises:

and encapsulating network performance parameters between the routing node and instance nodes corresponding to the computing power service instances indicated by the segment identifiers into the computing power request message.

21. The method of claim 20, wherein the power request message includes a segment routing header SRH, and wherein an extended type length value TLV field in the SRH includes the network performance parameter.

22. The method of any of claims 18 to 21, wherein prior to encapsulating the plurality of segment identifications corresponding to the identifications of the target computing force services in the initial request message, the method further comprises:

screening a plurality of segment identifiers meeting packaging conditions from a plurality of alternative segment identifiers corresponding to the identifiers of the target computing service;

wherein the packaging conditions include: the network performance between the routing node and the instance node corresponding to the computing power service instance indicated by the segment identification meets the communication condition.

23. The method of any of claims 18 to 22, wherein the plurality of segment identifiers are arranged in order of network performance from high to low.

24. The method according to any one of claims 15 to 23, wherein after forwarding the power request message according to the segment list, the method further comprises:

receiving an algorithm force response message sent by a target node, wherein the algorithm force response message comprises target segment identifiers of target algorithm force service examples corresponding to the target node, and the target segment identifiers belong to the plurality of segment identifiers;

and sending a power calculation task message to the target node, wherein a destination address field of the power calculation task message comprises the target segment identification.

25. The method of claim 24, wherein the destination segment identification is included in a source address field or an extension header of the power response message.

26. The method of claim 24 or 25, wherein the identity of the target computing force service is a common prefix of the plurality of segment identities; before sending the power calculation task message to the target node, the method further comprises:

based on the prefix of the target segment identifier being matched with the identifier of the target computing power service, the corresponding relation between the target computing power service and the target segment identifier is recorded.

27. The method of any of claims 15 to 26, wherein the segment list includes an identification of the target computing service.

28. A computing power request method, which is characterized by being applied to a computing power request system, wherein the computing power request system comprises a request node and a plurality of instance nodes corresponding to a plurality of computing power service instances, and each computing power service instance in the plurality of computing power service instances is used for providing a target computing power service; the method comprises the following steps:

the request node generates a calculation force request message, wherein a segment list of the calculation force request message comprises a plurality of segment identifiers, and the segment identifiers respectively indicate the plurality of calculation force service instances;

the request node forwards the calculation request message according to the segment list;

a first node in the plurality of instance nodes receives the computing force request message, and the first node corresponds to a first computing force service instance in the plurality of computing force service instances;

the first node determines whether to provide the target computing power service to a client device by the first computing power service instance based on whether the first computing power service instance satisfies a service condition.

29. The method of claim 28, wherein the first node determining whether to provide the target computing force service to a client device by the first computing force service instance based on whether the first computing force service instance satisfies a service condition comprises:

the first node determines that the first computing power service instance provides the target computing power service to the client device based on the first computing power service instance meeting the service condition, and sends a computing power response message to the client device; or,

the first node determines that the target computing power service is not provided to the client device by the first computing power service instance based on the first computing power service instance not meeting the service condition, and forwards the computing power request message to a second node of the plurality of instance nodes according to the segment list.

30. The method of claim 29, wherein the computing force response message includes a first segment identification indicating the first computing force service instance; the method further comprises the steps of:

the request node sends a power calculation task message to the first node, and a destination address field of the power calculation task message comprises the first section identifier.

31. A method according to any of claims 28 to 30, wherein the requesting node is the client device; before the requesting node generates the calculation force request message, the method further comprises:

the client device sends a domain name request for the target computing service to a Domain Name System (DNS) server;

the client device receives the plurality of segment identifications corresponding to the target computing power service sent by the DNS server.

32. A method according to any of claims 28 to 30, wherein the requesting node is a routing node connected to the client device; before the requesting node generates the calculation force request message, the method further comprises:

the routing node receives an initial request message sent by the client device, wherein the initial request message comprises an identifier of the target computing power service;

the routing node determines the plurality of segment identifications corresponding to identifications of the target computing power service.

33. A computing force requesting device, the computing force requesting device comprising: a memory, a processor and a computer program stored on the memory and capable of running on the processor, the processor implementing the method of any one of claims 1 to 27 when the computer program is executed.

34. A computer readable storage medium having instructions stored therein which, when executed on a processor, cause the processor to perform the method of any of claims 1 to 27.

35. A computer program product comprising instructions for execution by a processor to implement the method of any one of claims 1 to 27.

36. A computing force request system, the computing force request system comprising: a plurality of instance nodes corresponding to a plurality of computing force service instances, each computing force service instance of the plurality of computing force service instances for providing a target computing force service;

at least one of the plurality of instance nodes is configured to implement the method of any of claims 1 to 14.

37. The system of claim 36, wherein each instance node in the plurality of instance nodes is a service node to which its corresponding computing power service instance belongs;

or each instance node in the plurality of instance nodes is a routing node directly connected with the service node to which the corresponding computing power service instance belongs.

38. The system of claim 36 or 37, wherein the system further comprises: a requesting node for implementing the method of any one of claims 15 to 27.

39. The system of claim 38, wherein the requesting node is a client device or the requesting node is a routing node connected to the client device.