WO2022218341A1

WO2022218341A1 - Data forwarding method and related apparatus

Info

Publication number: WO2022218341A1
Application number: PCT/CN2022/086603
Authority: WO
Inventors: 顾炯炯; 苗勇
Original assignee: 华为云计算技术有限公司
Priority date: 2021-04-16
Filing date: 2022-04-13
Publication date: 2022-10-20
Also published as: CN115225631A

Abstract

A data forwarding method and a related apparatus, which are used for improving the degree of coverage of a network acceleration service. In an embodiment of the present application, a communication system comprises a central controller and a plurality of acceleration nodes, wherein the plurality of acceleration nodes comprise a first acceleration node and a second acceleration node, a deployment environment of the central controller belongs to a first cloud service provider, and a deployment environment of the plurality of acceleration nodes belongs to a second cloud service provider, an application service provider, or a telecommunication operator. The method comprises: a first acceleration node receiving a data request from a first terminal, wherein the data request is used for accessing a destination end; the first acceleration node acquiring a target path, wherein the target path is from a routing table entry generated by a central controller; and the first acceleration node sending the data request to a next-hop acceleration node according to the target path until the data request is forwarded to a second acceleration node, wherein the second acceleration node is used for forwarding the data request to the destination end. A plurality of forwarding nodes are flexibly deployed, thus improving the degree of coverage of a data forwarding network.

Description

A data forwarding method and related device

This application claims the priority of the Chinese patent application with the application number 202110411432.1 and the invention titled "A data forwarding method and related device" filed with the China Patent Office on April 16, 2021, the entire contents of which are incorporated herein by reference middle.

technical field

The present application relates to the technical field of computer networks, and in particular, to a data forwarding method and related devices.

Background technique

Based on various advantages of cloud services, more and more business applications choose cloud services provided by cloud vendors in order to provide high-quality services for business applications. The Global Accelerator (GA) network acceleration service emerges as the times require, thereby providing high-performance network acceleration services for global users.

Referring to Figure 1A, a cloud vendor builds a point of presence (POP) around the world, and the POP point is interconnected with the cloud vendor's private line network. The terminal is connected to the nearest POP point, for example, the terminal of the Asia-Pacific user is connected to POP1. The POP1 point introduces the data flow from the terminal to the private line network, and the private line network is connected to the cloud area, so that the terminal can quickly forward the data to be forwarded to the cloud area through the private line network deployed by the cloud manufacturer.

However, the GA service is completely dependent on the construction and distribution of POP nodes and physical private line networks deployed by cloud vendors, and the GA service capability is limited.

SUMMARY OF THE INVENTION

Embodiments of the present application provide a data forwarding method and a related device, which are used to improve the coverage of a network acceleration service.

In a first aspect, an embodiment of the present application provides a data forwarding method, which is applied to a first acceleration node in a communication system. The communication system includes a central controller and multiple acceleration nodes, and the multiple acceleration nodes belong to an overlay network, and the multiple acceleration nodes belong to an overlay network. The nodes include a first acceleration node and a second acceleration node, wherein the deployment environment of the central controller belongs to the first cloud service provider, and the deployment environment of the multiple acceleration nodes belongs to the second cloud service provider, application service provider or telecom operator During the operation of multiple acceleration nodes, they are controlled by the central controller, that is, controlled by the first cloud service provider; during the data forwarding process, the first acceleration node receives the data request from the first terminal, and the data request Used to access the destination; the first acceleration node obtains the target path, and the target path comes from the routing table entry generated by the central controller; the first acceleration node sends a data request to the next-hop acceleration node according to the target path, until the data request is forwarded to The second acceleration node is configured to forward the data request to the destination end. In this embodiment, multiple forwarding nodes can be flexibly deployed in a deployment environment provided by a second cloud service provider, an application service provider, or a telecom operator, thereby ensuring that acceleration nodes can be parasitized everywhere in the world, and all Any acceleration node among the acceleration nodes can be used as an access node for a terminal to access the network, and each acceleration node can be used as a transmission node in the target path. After the first acceleration node obtains the data request from the first terminal, , the first acceleration node sends the data request from the terminal to the overlay network according to the target path, until the data request is transmitted to the destination acceleration node (second acceleration node) of the target path, and the second acceleration node transmits the data request to the destination end, In this way, users around the world can truly enjoy the network acceleration service.

In an optional implementation manner, the routing table entry includes a source routing table and a location routing table; the method further includes:

The first acceleration node receives the source routing table and the location routing table sent by the central controller, the source routing table includes a path from the source acceleration node to the destination acceleration node, and the location routing table includes the first IP address and the second acceleration node. The corresponding relationship, The first IP address is the IP address of the destination end where the user applies for the network acceleration service; the acquisition of the target path by the first acceleration node may specifically include: when the destination address of the data request is the first IP address, the first acceleration node queries the location routing table , determine the second acceleration node corresponding to the first IP address, and the second acceleration node is the destination acceleration node; when the first acceleration node is the source acceleration node, the first acceleration node queries the source routing table according to the second acceleration node, and determines from The target path from the first acceleration node to the second acceleration node. In this embodiment, the first acceleration node can determine the second acceleration node according to the location routing table. When the first acceleration node is the source acceleration node and the second acceleration node is the destination acceleration node, the first acceleration node can determine the second acceleration node according to the source routing table. A target path from the first acceleration node to the second acceleration node is determined, and the first acceleration node forwards to the next-hop acceleration node according to the target path until forwarding to the second acceleration node.

In an optional implementation manner, before the first acceleration node receives the source routing table sent by the central controller, the method further includes: the first acceleration node measures the link status between the first acceleration node and neighboring acceleration nodes, The link state information is obtained; the first acceleration node sends the link state information to the central controller, and the link state information is used by the central controller to generate a source routing table. In this embodiment, the acceleration node is also used to measure the link status with the neighboring acceleration node, so that the central controller can generate a source routing table according to the link status, and the path in the source routing table is obtained according to the link status , so that the network acceleration service has a higher quality of service.

In an optional implementation manner, an SDK plug-in is configured in the first terminal, the address of the central controller is preset in the SDK plug-in, and the first acceleration node receiving the data request from the first terminal may include: the first acceleration node The SDK package data is received from the first terminal through the SDK tunnel. The SDK package data is the data after the data request is packaged. The destination address in the header of the SDK package data is the IP address of the first acceleration node, and the source address in the header is the IP address of the first terminal. In this embodiment, an SDK plug-in is configured in the terminal, the terminal can access an acceleration node nearby, and the overlay network performs accelerated forwarding for the data request of the first terminal, and the application scenarios are wide.

In an optional implementation manner, the deployment environment of the first acceleration node is a first network device, the first network device is used to receive an access control list ACL policy instruction, and the ACL policy instruction is used to trigger the first network device to convert the destination address The data of the first IP address is guided to the first acceleration node; the first acceleration node receiving the data request from the first terminal may include: the first acceleration node receiving the data from the first terminal guided by the first network device according to the ACL policy instruction ask. This embodiment is applicable to a scenario where a terminal accesses a network through a first network device (eg, MEC or OLT) (eg, a scenario where a home bandwidth accesses the network). The first network device not only serves as a network access device of the first terminal, the first acceleration node deployed in the first network device also serves as a source acceleration node in the overlay network, and the overlay network access points of the terminal are abundant.

In an optional implementation manner, the deployment environment of the first acceleration node is a device in a local area network, the first terminal is a terminal in the local area network, and receiving the data request from the first terminal by the first acceleration node may include: the first acceleration node The node receives the data request from the first terminal through the local area network. In this embodiment, the acceleration node is embedded and deployed in a local area network (such as an enterprise intranet), and the second network device provides a private network AIP for the first acceleration node, so that terminal devices in the local area network can access the overlay network through the acceleration node . The overlay network provides network acceleration services for terminal devices in the local area network, which solves the problem of expensive public network AIP resources allocated by operators.

In an optional implementation manner, the destination end is a cloud area, or the destination end is a second terminal or a server. In this embodiment, the overlay network not only supports a scenario in which a terminal accesses a cloud region, but also supports a scenario in which a terminal accesses between terminals, which is universal.

In an optional implementation, the deployment environment includes a cloud area, POP, edge cloud, OLT, or MEC. In this embodiment, the acceleration node can be flexibly deployed in various deployment environments, so that the overlay network can truly cover the global area.

In a second aspect, an embodiment of the present application provides a data forwarding method, which is applied to a central controller in a communication system. The communication system includes a central controller and multiple acceleration nodes deployed in various deployment environments. The multiple acceleration nodes It includes a first acceleration node and a second acceleration node, and the method includes: the central controller obtains link state information sent by the acceleration node; the central controller generates a source routing table according to the link state information, and the source routing table contains the information from the source acceleration node. The path to the destination acceleration node; the central controller obtains the first IP address of the destination end where the user applies for the network acceleration service; the central controller generates a location routing table, and the location routing table includes the correspondence between the first IP address and the second acceleration node; the center The controller sends the location routing table and the source routing table corresponding to the first acceleration node to the first acceleration node, where the location routing table is used to guide the first acceleration node to determine the second acceleration node according to the first IP address, and the first IP address is from the first acceleration node. The destination address of a data request of a terminal, the source routing table is used by the first acceleration node to obtain the target path, the target path is used to guide the data request to be forwarded to the second acceleration node, and the second acceleration node is used to forward the data to the destination. In this embodiment, the central controller generates a source routing table, the source routing table is used to indicate paths in multiple acceleration nodes, the central controller generates a location routing table, and the location routing table is used to indicate the destination acceleration node, so that the source acceleration node It can query the destination acceleration node according to the location routing table, and then query the target path according to the source routing table, so as to realize the accelerated forwarding of data requests by the acceleration node in the overlay network.

In an optional implementation manner, the generation of the location routing table by the central controller may include: the central controller determines the second acceleration node according to the first IP address of the destination terminal; the central controller establishes a relationship between the first IP address and the second acceleration node Corresponding relationship; the central controller generates a location routing table according to the corresponding relationship.

In an optional implementation manner, when the destination terminal is a cloud area, the central controller determining the second acceleration node according to the first IP address of the destination terminal may include: the central controller determining the acceleration node deployed in the cloud area according to the first IP address The second acceleration node. In this embodiment, the acceleration node can be flexibly deployed in the cloud area. When the destination is in the cloud area, the central controller directly determines the second acceleration node deployed in the cloud area according to the first IP address, so that the second acceleration node can be deployed in the cloud area. The node forwards the data request to the destination within the cloud area, reducing the transmission distance from the acceleration node to the destination.

In an optional implementation manner, when the destination terminal is a terminal or a server, the central controller determining the second acceleration node according to the first IP address of the destination terminal may include: the central controller querying an IP address library according to the first IP address, Determine the physical location of the destination; the central controller determines the second acceleration node closest to the physical location, thereby minimizing the transmission distance from the second acceleration node to the destination.

In an optional implementation manner, an SDK plug-in is configured in the first terminal, and address information of the central controller is preset in the SDK plug-in; the method further includes: the central central controller receives a request sent by the first terminal; The central controller feeds back the IP address of the first acceleration node to the first terminal according to the request, and the IP address of the first acceleration node is used by the first terminal to send a data request to the first acceleration node by using the SDK tunnel. An SDK plug-in is configured in the first terminal, and the address of the central controller is preset in the SDK plug-in. The first terminal accesses the central controller through the SDK plug-in, that is, the central central controller receives a request sent by the first terminal, the request carries the IP address of the first terminal, and the central controller queries the IP address database according to the IP address of the first terminal, and determines the physical location of the first terminal, and determine the acceleration node (that is, the first acceleration node) closest to the physical location according to the physical location of the first terminal, and the central controller feeds back the IP address of the first acceleration node to the first terminal, overlay The network performs accelerated forwarding for the data request of the first terminal, and has a wide range of application scenarios.

In an optional implementation manner, the method further includes: the central controller sends traffic diversion information to the network management system, the traffic diversion information includes IP information of the destination end, and the traffic diversion information is used to trigger the network management system to send the ACL to the first network device The policy instruction, the second acceleration node is an acceleration node deployed in the first network device, and the ACL policy instruction is used to trigger the first network device to direct the data request from the first terminal to the first acceleration node. In this embodiment, the central controller and the first network device cooperate to guide the data request of the first terminal to the first acceleration node. It is applicable to the scenario where the terminal accesses the network through the first network device (eg, MEC or OLT) (eg, the scenario where the home bandwidth accesses the network). The first network device not only serves as a network access device of the first terminal, the first acceleration node deployed in the first network device also serves as a source acceleration node in the overlay network, and the overlay network access points of the terminal are abundant.

In an optional implementation manner, the method further includes: the central controller obtains a mode parameter, where the mode parameter includes a first mode and a second mode, wherein the first mode is used to indicate that the destination of the network acceleration service is the cloud area, and the second mode is used to indicate that the destination of the network acceleration service is the second terminal or server. In this embodiment, the overlay network not only supports a scenario in which a terminal accesses a cloud region, but also supports a scenario in which a terminal accesses between terminals, which is universal.

In a third aspect, an embodiment of the present application provides an acceleration node, which is included in a communication system. The communication system includes a central controller and multiple acceleration nodes, and the multiple acceleration nodes include a first acceleration node and a second acceleration node, wherein the center The deployment environment of the controller belongs to the first cloud service provider, and the deployment environment of the multiple acceleration nodes belongs to the second cloud service provider, an application service provider or a telecom operator; the first acceleration node includes:

a forwarding module, configured to receive a data request from the first terminal, and the data request is used to access the destination terminal;

The control module is used to obtain the target path, and the target path comes from the routing table entry generated by the central controller;

The forwarding module is configured to send a data request to the next-hop acceleration node according to the target path until the data request is forwarded to the second acceleration node, and the second acceleration node is configured to forward the data request to the destination.

In an optional implementation manner, the routing table entry includes a source routing table and a location routing table;

The control module is also used to receive the source routing table and the location routing table sent by the central controller, the source routing table includes a path from the source acceleration node to the destination acceleration node, and the location routing table includes the first IP address and the second acceleration node. Correspondence relationship, the first IP address is the IP address of the destination end where the user applies for the network acceleration service;

When the destination address of the data request is the first IP address, the first acceleration node queries the location routing table to determine the second acceleration node corresponding to the first IP address, and the second acceleration node is the destination acceleration node;

When the first acceleration node is the source acceleration node, the first acceleration node queries the source routing table according to the second acceleration node, and determines the target path from the first acceleration node to the second acceleration node.

In an optional implementation manner, the forwarding module is further configured to measure the link state between the first acceleration node and the neighbor acceleration node to obtain link state information;

The control module is further configured to send the link state information obtained by the forwarding module to the central controller, where the link state information is used by the central controller to generate a source routing table.

In an optional implementation manner, an SDK plug-in is configured in the first terminal, and the address of the central controller is preset in the SDK plug-in;

The forwarding module is further configured to receive the SDK encapsulation data from the first terminal through the SDK tunnel, the SDK encapsulation data is the data after encapsulating the data request, and the destination address in the header of the SDK encapsulation data is the IP address of the first acceleration node, The source address in the header is the IP address of the first terminal.

In an optional implementation manner, the deployment environment of the first acceleration node is a first network device, the first network device is used to receive an access control list ACL policy instruction, and the ACL policy instruction is used to trigger the first network device to convert the destination address The data of the first IP address is directed to the first acceleration node;

The forwarding module is further configured to receive a data request from the first terminal guided by the first network device according to the ACL policy instruction.

In an optional implementation manner, the deployment environment of the first acceleration node is a device in a local area network, and the first terminal is a terminal in the local area network;

The forwarding module is further configured to receive a data request from the first terminal through the local area network.

In a fourth aspect, an embodiment of the present application provides a central controller, including:

The transceiver module is used to obtain the link status information sent by the acceleration node;

a processing module, configured to generate a source routing table according to the link state information acquired by the transceiver module, where the source routing table includes a path from the source acceleration node to the destination acceleration node;

The transceiver module is used to obtain the first IP address of the destination end where the user applies for the network acceleration service;

The processing module is also used to generate a location routing table, where the location routing table includes the correspondence between the first IP address and the second acceleration node;

The transceiver module is further configured to send the location routing table and the source routing table corresponding to the first acceleration node to the first acceleration node, where the location routing table is used to guide the first acceleration node to determine the second acceleration node according to the first IP address, the first IP address The address is the destination address of the data request from the first terminal. The source routing table is used by the first acceleration node to obtain the target path, and the target path is used to guide the data request to be forwarded to the second acceleration node, and the second acceleration node is used to forward the data to destination.

In an optional implementation manner, the processing module is further specifically configured to: determine the second acceleration node according to the first IP address of the destination end; establish a correspondence between the first IP address and the second acceleration node; generate a location route according to the correspondence surface.

In an optional implementation manner, when the destination terminal is the cloud area, the processing module is further configured to determine the second acceleration node deployed in the cloud area according to the first IP address.

In an optional implementation manner, when the destination terminal is a terminal or a server, the processing module is further configured to query the IP address database according to the first IP address to determine the physical location of the destination terminal; and determine the second accelerator closest to the physical location. node.

In an optional implementation manner, an SDK plug-in is configured in the first terminal, and the address information of the central controller is preset in the SDK plug-in; the transceiver module is further configured to receive a request sent by the first terminal; A terminal feeds back the IP address of the first acceleration node, and the IP address of the first acceleration node is used by the first terminal to send a data request to the first acceleration node by using the SDK tunnel.

In an optional implementation manner, the transceiver module is further configured to send traffic diversion information to the network management system, the traffic diversion information includes IP information of the destination end, and the traffic diversion information is used to trigger the network management system to send an ACL policy instruction to the first network device, The second acceleration node is an acceleration node deployed in the first network device, and the ACL policy instruction is used to trigger the first network device to direct the data request from the first terminal to the first acceleration node.

In an optional implementation manner, the transceiver module is further configured to obtain a mode parameter, where the mode parameter includes a first mode and a second mode, wherein the first mode is used to indicate that the destination end of the network acceleration service is a cloud area, and the first mode is used to indicate that the destination end of the network acceleration service is a cloud area. The second mode is used to indicate that the destination of the network acceleration service is the second terminal or server.

In a fifth aspect, an embodiment of the present application provides a communication system, including a plurality of acceleration nodes according to the third aspect and a central controller according to the fourth aspect, wherein the deployment of the central controller The environment belongs to the first cloud service provider, and the deployment environment of the multiple acceleration nodes belongs to the second cloud service provider, an application service provider or a telecom operator.

In a sixth aspect, an embodiment of the present application provides a central controller, including a processor, the processor is coupled to at least one memory, and the processor is configured to read a computer program stored in the at least one memory, so that all The central controller executes the method described in any one of the above second aspects.

In a seventh aspect, an embodiment of the present application provides a computer program product, the computer program product includes computer program code, and when the computer program code is executed by a computer, enables the computer to implement any one of the above-mentioned first aspects. or, causing a computer to implement the method described in any one of the second aspects above.

In an eighth aspect, an embodiment of the present application provides a computer-readable storage medium for storing a computer program or instruction, and when the computer program or instruction is executed, the computer executes the method described in any one of the first aspect above; Alternatively, the computer is caused to implement the method described in any one of the above second aspects.

Description of drawings

1A and 1B are schematic diagrams of scenarios of a network acceleration system in a traditional method;

FIG. 2 is a schematic diagram of a scenario of a communication system in an embodiment of the present application;

3 is a schematic diagram of overlay and underlay in an embodiment of the application;

FIG. 4 is a schematic structural diagram of a communication system in an embodiment of the application;

FIG. 5 is a schematic flowchart of steps of an embodiment of a data forwarding method in an embodiment of the present application;

6 is a schematic diagram of a scenario in which a first terminal accesses a first acceleration node in an embodiment of the present application;

7 is a schematic diagram of a scenario in which a central controller and a first network device cooperate to guide a data request to a first acceleration node in an embodiment of the present application;

FIG. 8 is a schematic diagram of two application modes for a business application to apply for a network acceleration service in an embodiment of the present application;

9 is a schematic diagram of a scenario of an application interface for a network acceleration service in an embodiment of the present application;

10 is a schematic diagram of a scenario in which a terminal accesses a cloud region in an embodiment of the application;

11 is a schematic diagram of a data format for overlay encapsulation based on UDP in an embodiment of the application;

12 is a schematic diagram of a scenario of data forwarding between a first terminal and a cloud region in an embodiment of the application;

13 is a schematic diagram of a scenario of data forwarding between a first terminal and a second terminal in an embodiment of the present application;

14 is a schematic diagram of an overlay tunnel encrypting and forwarding data in an embodiment of the present application;

FIG. 15 is a schematic structural diagram of an embodiment of an acceleration node in an embodiment of the present application;

16 is a schematic diagram of the architecture of a virtual machine in an embodiment of the present application;

17 is a schematic structural diagram of an embodiment of a central controller in an embodiment of the application;

FIG. 18 is a schematic structural diagram of another embodiment of the central controller in the embodiment of the present application.

Detailed ways

The technical solutions in the embodiments of the present application will be described below with reference to the accompanying drawings in the embodiments of the present application. The terms "first", "second", etc. in the description and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence.

Referring to FIG. 1B , the network acceleration system in the conventional method includes a DNS server, a controller, a plurality of POP points, and an IP dedicated line network connected to the POP points. Each POP point is configured with at least one anycast IP (anycast IP, AIP) address. When a business application applies for the acceleration service, the controller generates a mapping relationship between AIP and an elastic IP address (EIP) (public network IP address). For example, business application A applies for a network acceleration service, the destination end of the network acceleration service is a cloud region (region), and the EIP of the cloud region is EIP1. After receiving the application from the service application A, the controller allocates an access AIP (eg, AIP1) to the service application. The controller maintains the identity of the service application A, the mapping relationship between EIP1 and AIP1 (as shown in Table 1 below), and delivers the mapping relationship to each POP synchronously. The controller sends the mapping relationship between EIP1 and AIP1 to the DNS server, and the DNS server is used to synchronously maintain the mapping relationship between the domain name, AIP and EIP. The mapping relationship between each AIP and EIP is shown in Table 1 below.

Table 1

业务应用Abusiness application A	EIP1EIP1	AIP1AIP1
业务应用Bbusiness application B	EIP2EIP2	AIP2AIP2
业务应用Cbusiness application C	EIP3EIP3	AIP3AIP3

The GA scheme in the traditional method includes two stages, the first stage: the stage of terminal access to the POP point. The second stage: the POP point accesses the cloud region (region) stage through the IP private line network.

Stage 1: First, the terminal sends the domain name of the resource to be accessed to the domain name system (DNS) server. Then, the DNS server feeds back the AIP (eg, AIP2) that has a mapping relationship with the domain name to the terminal. For example, the number of POP points configured with AIP2 is 3. Afterwards, the terminal accesses the "nearest route" POP point (such as POP2) through the underlay. The terminal sends a data packet to the POP2, the source address in the packet header of the data packet is the IP address of the terminal (such as IPA), and the destination address is AIP2. Finally, POP2 modifies the destination address in the original data packet from the terminal to EIP2 through the mapping relationship shown in Table 1 above (AIP2 and EIP2 have a mapping relationship) to obtain the target data packet. The source address of the target data packet is IPA, The destination address is EIP2. POP2 sends the target data packet to the IP private line network. It should be noted that the real destination address (EIP2) in the original data packet is lost, and the lost EIP2 is restored by the POP2 node according to the mapping relationship between EIP1 and AIP1.

The second stage: The POP point is interconnected with the IP private line network. The POP point introduces the data flow from the terminal to the IP private line network, and the IP private line network forwards the data flow to the cloud region, so that the terminal can accelerate the access to the cloud region.

The GA network acceleration service method in the traditional method has at least the following problems.

1. GA acceleration service depends on the investment and construction of cloud vendors. At present, there are only dozens of POP points in the world. Cloud manufacturers have not invested in building POP points in some regions (or certain countries and regions) around the world, and the global coverage is not enough, which leads to the lack of POP points. Terminals in the region cannot access the IP private line network and cannot enjoy the network acceleration services provided by cloud vendors.

2. The EIP of a service application consumes one AIP address mapping, which leads to the consumption of a large amount of AIP. And adding anycast IP addresses to the routing network can be done by using routing protocols (such as border gateway protocol (BGP), and an AIP needs to be published to the Internet at multiple POP points BGP multi-source, and distributed across operators AIP is difficult, that is, different regions (such as China and India) have different operators who allocate the same AIP, and it is difficult for different operators to issue the same AIP project.

3. Only supports the scenario where the terminal accesses the region.

In view of the above problems, an embodiment of the present application provides a data forwarding method, which is based on an overlay overlay network on an underlay, and implements data forwarding through an overlay overlay network. The overlay network includes a central controller and a large number of acceleration nodes deployed in various deployment environments. For example, a large number of forwarding nodes are flexibly deployed on edge clouds, POP points, cloud regions, OLT and MEC devices around the world. The overlay network can truly cover the global area. The overlay network in this application is used to realize the forwarding of service application data, and the overlay is also called an application delivery network (application delivery network, ADN).

Please refer to FIG. 2 , which is a schematic diagram of a scenario of a communication system. The communication system includes a central controller 201 and a plurality of acceleration nodes 202 . The central controller 201 is used to manage and control all the acceleration nodes 202. Taking the first acceleration node among the multiple acceleration nodes 202 as an example, the forwarding of the data from the terminal by the first acceleration node will be described. The first acceleration node receives the routing table entry sent by the central controller 201, and the routing table entry serves as the basis for the first acceleration node to forward data. The first acceleration node receives a data request from the first terminal, and the data request is used to access the destination; first, the first acceleration node obtains the target path, and the target path comes from the routing table entry generated by the central controller 201; An acceleration node sends a data request to the next-hop acceleration node according to the target path, until the data request is forwarded to the second acceleration node, and the second acceleration node is used for forwarding the data request to the destination.

In the embodiment of the present application, a large number of acceleration nodes are flexibly deployed in the overlay network, and any acceleration node among all the acceleration nodes can be used as the access acceleration node of the terminal, so that terminals around the world can access a nearby acceleration node. on the acceleration node. After the first acceleration node obtains the data request of the first terminal, the first acceleration node sends the data request from the terminal to the overlay network according to the target path, until the data request is transmitted to the destination acceleration node (second acceleration node) of the target path, The second acceleration node then transmits the data request to the destination end, so that users of business applications worldwide can truly enjoy the network acceleration service. And compared with the GA scheme in the traditional method, any acceleration node among all the acceleration nodes can be used as a terminal to access the access node of the overlay network, and each acceleration node can be used as a transmission node in the target path. The services provided by each acceleration node are shared by all destinations, and there is no need for GA in the traditional method. A business application needs to spend one AIP address mapping, and the project deployment is easy to implement.

For a better understanding of the present application, firstly, the words involved in the present application will be exemplified.

The central controller is used to control all the acceleration nodes, obtain the link status reported by the acceleration nodes, generate the source routing table according to the link status between the acceleration nodes, and also generate the location routing table, and connect the source routing table and the location routing table. The table is delivered to each acceleration node. The central controller can be a virtual server deployed on the cloud side.

Accelerate node is used to realize data forwarding function and link state measurement function. Acceleration nodes are deployed in virtual machines or containers provided by the deployment environment. The acceleration node includes a local controller and at least one forward node (forward node or compass). The local controller is used to control the compass to perform link state measurement (or also referred to as "QoS measurement") between accelerated nodes. Compass is mainly responsible for the traffic forwarding function of the data plane. The compass may be a forwarding module that implements the forwarding function through software.

The deployment environment of the acceleration node, which is used to assign a "host" and a public IP address to the acceleration node. Deployment environments include but are not limited to edge cloud, optical line terminal (OLT), multi-access edge computing (MEC), POP, cloud region (region), content delivery network (content delivery network) network, CDN), etc. It can be understood that the deployment environment only needs to provide a virtual machine (or container) and a public network IP address, and then the acceleration node can be deployed. This requirement is not harsh or customized. For example, third-party CDNs, edge clouds, OLT or MEC devices can easily provide virtual machines (or containers) and public IP addresses, so that acceleration nodes can be parasitized into the deployment environment. Due to the flexibility in the deployment of acceleration nodes, it is ensured that the acceleration nodes can be parasitized everywhere in the world, and the overlay network can cover a wider range of global areas. For another example, the flexible deployment of acceleration nodes is also reflected in the fact that the acceleration nodes can be deployed in the cloud region. When the acceleration nodes are deployed in the cloud region, the cloud region only needs to provide virtual machines (or containers) and public network IPs for the acceleration nodes to meet the deployment conditions. . Therefore, the deployment of acceleration nodes can meet all cloud types such as partner cloud, third-party cloud, partner cloud, and hybrid cloud, and the application business scope of acceleration is wider. For example, "video service provider A" can directly deploy acceleration nodes to the third-party edge cloud built by "video service provider A" to provide acceleration services for "video service provider A"'s services.

Additionally, multiple deployment environments in a communication system may belong to different providers. The provider can be a cloud service provider, an application service provider (such as "instant messaging service" provider A), or a telecom operator (such as China Mobile, China Unicom, and Telecom). The deployment environment of the central controller belongs to the first cloud service provider (such as cloud service provider A), and the deployment environment of multiple acceleration nodes may belong to the second cloud service provider (such as cloud service provider B, cloud service provider C) ), application service provider or telecom operator. The second cloud service provider, application service provider or telecom operator provides a deployment environment of the acceleration node on its own hardware facilities. The deployment environment here is a virtual environment for the first cloud service provider to apply for computing resources (eg, virtual machines, containers). Or apply for computing resources on the deployment environment provided by telecom operators and run acceleration nodes on the computing resources. During the operation of the acceleration node, it is controlled by the central controller, that is, controlled by the first cloud service provider.

Service application, the user of the service traffic forwarding service provided by the overlay network. For example, including but not limited to real time communication (real time communication, RTC) service, cloud video service, game service, video service and other application platforms.

Terminals, including but not limited to server terminals, mobile phone (mobile phone), tablet computer (Pad), personal computer (personal computer, PC), virtual reality (virtual reality, VR) terminal, augmented reality (augmented reality, AR) terminal, Terminals in industrial control, in-vehicle terminals, terminals in self-driving, terminals in assisted driving, terminals in remote medical, terminals in smart grid, Terminals in transportation safety, terminals in smart cities, terminals in smart homes, etc.

The destination end may be a cloud area, or may be a second terminal or a server.

The source routing table and the location routing table are exemplified.

The source routing table is used to indicate the optimal path from the source acceleration node to the destination acceleration node. Among them, "source acceleration node" and "destination acceleration node" are both acceleration nodes in the above overlay network. " and "Destination Acceleration Node". For example, an acceleration node that receives data from a terminal is a "source acceleration node". The acceleration node that sends data to the destination is the "destination acceleration node". The source acceleration node can traverse every acceleration node in all forwarding nodes, and the destination acceleration node can traverse every acceleration node in all forwarding nodes.

The location routing table includes the correspondence between the first IP address and the second acceleration node. Wherein, the first IP address is the IP address of the destination end where the user (such as a business application) applies to the central controller in advance for the acceleration service. The second acceleration node is a forwarding node determined by the central controller according to the IP address of the destination end. The second acceleration node is the acceleration node closest to the destination among all the acceleration nodes. For example, when the destination is a cloud area, the second acceleration node may be an acceleration node deployed in the cloud area. For another example, when the destination is a terminal (or server), the central controller queries the IP address database to determine the physical location of the terminal (or server), and the central controller determines the acceleration node (ie, the closest acceleration node to the physical location of the terminal (or server)). second acceleration node). It should be noted that, in order to distinguish the source acceleration node and the destination acceleration node, any one of the multiple source acceleration nodes is called the "first acceleration node", and any one of the multiple destination acceleration nodes is called the "first acceleration node". "Second Acceleration Node".

Link status, the acceleration node measures the quality of service (QoS) of its neighbor acceleration nodes to obtain link status information. The link state information includes the link state of the acceleration node to each neighbor acceleration node. It can be understood that the link state can be described by a QoS value, wherein the performance indicators of the QoS include packet loss rate, delay and jitter, and the like. Exemplarily, when the acceleration node performs QoS measurement on the link between each neighbor acceleration node, the acceleration node will continuously send q probe packets (q is an integer greater than or equal to 2) to its neighbor acceleration node. The replies of the q probe packets are used to calculate the transmission delay, jitter, and packet loss rate. Optionally, the acceleration node performs a weighted average of transmission delay, jitter and packet loss rate, and uses the weighted average value to describe the link state between the acceleration node and the neighbor acceleration node. In this application, the measurement of the "link status" of the acceleration node of its neighbor acceleration nodes may also be referred to as "QoS measurement".

A description of overlay networks and underlay networks.

The underlay network refers to the physical network, which consists of physical devices and physical links. For example, common physical devices include switches, routers, and firewalls. These physical devices are connected through specific links to form a traditional physical network.

An overlay network is a computer network that can be built on top of an underlay. Nodes (ie, forwarding nodes) in an overlay network can be considered to be connected by virtual or logical links, where each link corresponds to a path. Please refer to FIG. 3 for understanding. The four nodes H, I, J, and K in FIG. 3 are logical nodes in the overlay network. Exemplarily, in the upper-layer overlay network, the direct connection between the H and J nodes, that is, there is only one hop at the application layer level, is mapped to the lower-layer underlay network, which may involve multiple relay forwarding routing devices, It is actually multi-hop routing. However, when developing applications based on the overlay network, it is not necessary to consider the connection of each physical node in the underlying underlay network, but only the interconnection between nodes in the deployed overlay. Nodes in the overlay network implement data forwarding at the overlay layer by encapsulating the source IP and destination IP mapped to the nodes in the underlay network.

Full-mash refers to a networking mode in which two nodes among N nodes are interconnected.

The data forwarding method in the present application will be exemplarily described below through specific embodiments. Please refer to FIG. 4 , which is a schematic structural diagram of a communication system in the present application.

First, the process of deploying an acceleration node in a deployment environment in the present application is described with an example. The operation and maintenance personnel in the following S30-S33 are the operation and maintenance personnel of the first cloud provider.

S30, the PC of the operation and maintenance personnel, in response to the first operation of the operation and maintenance personnel, applies to the deployment environment such as edge cloud, POP, cloud region, etc. to allocate a virtual machine (or container) and a public network IP. Wherein, an operation interface for applying for a virtual machine (or container) and a public network IP for the deployment environment is installed in the PC.

S31. In response to the second operation of the operation and maintenance personnel, the PC of the operation and maintenance personnel logs in to the account of the deployment center, uses the deployment center to automatically upload to the virtual machine (or container), and automatically installs acceleration node software in batches. Among them, the deployment center is a cloud center tool for automated batch deployment of accelerated nodes.

S32. After the installation of the acceleration node is completed, actively initiate handshake authentication to the central controller.

S33: The central controller handshakes and communicates with the forwarding node, and the central controller receives the registration request sent by each forwarding node. The registration request includes but is not limited to the ID of the deployment environment of the forwarding node, the public IP address of the forwarding node, the physical location of the deployment environment of the forwarding node, and the like. The process of registering the forwarding node by the central controller can be understood as a process of storing the relevant information of the forwarding node by the central controller. The central controller obtains the relevant information of each forwarding node, and can further manage each forwarding node.

The above-mentioned steps S30 to S33 are the deployment process of the acceleration node. If the forwarding node has been registered to the central controller, and the registered acceleration node has not been deleted, or no other acceleration node has been added, it is not necessary to perform step S30 every time. - Step S33. Steps S30 to S33 are optional steps, and step 501 is directly executed.

Then, see Figure 5. The process of realizing data forwarding in this application is described.

Step 501: The central controller acquires link state information between the acceleration node and the neighboring acceleration nodes.

The central controller receives the link state information sent by each acceleration node, and the link state information includes the link state between the acceleration node and each neighbor acceleration node.

Exemplarily, the central controller sends measurement tasks to the local controllers in each acceleration node. All acceleration nodes are fully interconnected (full-mash), and the acceleration node performs QoS measurement on the links between its neighbor acceleration nodes. Each acceleration node will collect link status information, the link status information includes the link status (described by QoS value) from the acceleration node to neighboring acceleration nodes and the link identifier corresponding to the link status (for example, the acceleration node Node A → Accelerate Node B). The acceleration node in the neighborhood of the acceleration node refers to a node connected to the acceleration node. Taking the full interconnection of all acceleration nodes as an example, the neighbor acceleration nodes of any acceleration node refer to all other acceleration nodes except the forwarding node among all the acceleration nodes. Optionally, the compass in the acceleration node performs QoS measurement periodically (for example, in seconds), collects link state information (represented by a QoS value), and stores the collected link state in the local controller. The local controller reports link status information to the central controller periodically (for example, in minutes).

Step 502: The central controller generates a source routing table according to the link state information. The source routing table is used to indicate paths in multiple acceleration nodes. The path is the path from the source acceleration node to the destination acceleration node.

S11. The central controller selects a path from the source acceleration node to the destination acceleration node among all the acceleration nodes according to the link state information and the topology structures of all the acceleration nodes.

The central controller receives the link status information reported by each acceleration node. The topology of all acceleration nodes is fully interconnected as an example, and the central controller determines multiple paths. For example, taking the acceleration node A as the source forwarding node, the central controller calculates the paths from the acceleration node A to other acceleration nodes (eg, the acceleration node B and the acceleration node C). Taking the acceleration node B as the source forwarding node, the central controller calculates the path from the acceleration node B to other acceleration nodes (eg, the acceleration node A and the acceleration node C).

S12. The central controller generates a source routing table corresponding to each acceleration node based on the path.

The source routing table includes a list of acceleration nodes experienced by the path and next-hop acceleration nodes. The next-hop acceleration nodes of each acceleration node may be different. Therefore, each forwarding node needs to correspond to a different source routing table.

Step 503: The central controller delivers the source routing table corresponding to the acceleration node to each acceleration node.

The acceleration node takes the first acceleration node and the second acceleration node as examples. For example, the central controller sends the source routing table A to the first acceleration node. The central controller sends the source routing table B to the second acceleration node.

Step 504: The central controller obtains the first IP address of the destination end of the user applying for the network acceleration service.

For example, the central controller receives a request for applying for a network acceleration service, where the request carries a service domain name. The central controller sends the service domain name to the DNS server, and the DNS server is used to resolve the service domain name to obtain IP information (eg, EIP1 ) of the cloud region, and the first IP address is EIP1 . The central controller obtains the IP information (ie EIP1) of the cloud region from the DNS server. For another example, where the request carries the IP address (eg IP1) of the terminal (or server), the first IP address is IP1.

Step 505: The central controller generates a location routing table, where the location routing table includes the correspondence between the first IP address and the second acceleration node.

First, the central controller determines the second acceleration node according to the first IP address of the destination. When the destination is the cloud region, that is, the first IP address is the EIP of the cloud region, the central controller determines the acceleration node (eg, the second acceleration node) deployed in the cloud region. Another example, when the destination is a terminal (or server), that is, when the first IP address is the IP address of the terminal (or server), the central controller searches the IP address library to determine the physical location of the destination, and the central controller determines the distance from the The second acceleration node with the closest physical location. Then, the central controller establishes the correspondence between the first IP address and the second acceleration node. Finally, the central controller generates a location routing table according to the corresponding relationship.

Step 506: The central controller sends the location routing table to all the acceleration nodes, and all the acceleration nodes include the first acceleration node; correspondingly, the first acceleration node receives the source routing table and the location routing table sent by the central controller.

The central controller calls the southbound data interface to deliver the location routing table to the local controllers of each acceleration node, and the local controller delivers the source routing table and the location routing table to Compass. It should be understood that the central controller sends the source routing table corresponding to each acceleration node to each acceleration node. Instead, send the same location routing table to all acceleration nodes. For example, the central controller sends the source routing table A and the location routing table to the first acceleration node.

For the convenience of description, the process of forwarding data by the first acceleration node among all the acceleration nodes is taken as an example for description, and the first acceleration node is any acceleration node among all the acceleration nodes.

Step 507: The first acceleration node receives the data request from the first terminal.

In the first implementation manner, the first terminal accesses the first acceleration node through a software development kit (SDK) tunnel, that is, the first acceleration node receives the SDK package data through the SDK tunnel, and the SDK package data is a request for data Encapsulated data.

Exemplarily, please refer to FIG. 6 , which is a schematic diagram of a scenario in which the first terminal accesses the first acceleration node. An SDK plug-in is configured in the first terminal, and the address of the central controller is preset in the SDK plug-in. The first terminal accesses the central controller through the SDK plug-in, that is, the central central controller receives a request sent by the first terminal, where the request carries the IP address of the first terminal. The central controller queries the IP address library according to the IP address of the first terminal, determines the physical location of the first terminal, and determines the acceleration node (ie, the first acceleration node) closest to the physical location according to the physical location of the first terminal, and the central controller controls the The controller feeds back the IP address (eg IP2) of the first acceleration node to the first terminal.

For example, if the destination end accessed by the first terminal is a cloud region (IP address is EIP1), the first terminal sends the raw data to be sent (also referred to as a "data request") to the first acceleration node. The destination address of the original data is EIP1, and the source address of the original data is IPA (ie, the IP address of the first terminal). The first terminal performs SDK encapsulation on the data request to obtain SDK encapsulation data. The destination address in the header (or "packet header") of the SDK encapsulated data is the IP address (eg IP2) of the first acceleration node, and the source address in the header is the IP address (eg IPA) of the first terminal.

In this embodiment, an SDK plug-in is configured in the terminal, the terminal can access an acceleration node nearby, and the overlay network performs accelerated forwarding for the data request of the first terminal, and the application scenarios are wide.

In the second implementation manner, the central controller and the first network device cooperate to guide the data request of the first terminal to the first acceleration node.

Please refer to FIG. 7 , which is a schematic diagram of a scenario in which the central controller and the first network device cooperate to guide the data request to the first acceleration node. The first acceleration node is an acceleration node deployed in the first network device. For example, the first network device may be an MEC or an OLT.

First, the central controller sends traffic diversion information to the network management system, and the network management system is used to manage network element equipment (such as MEC or OLT, etc.). The traffic diversion information includes the IP address (ie, the first IP address) of the destination end. The network management device generates an access control list (access control list, ACL) policy instruction according to the traffic diversion information, and sends the ACL policy instruction to the first network device, where the ACL policy instruction is used to trigger the first network device to assign the destination address to the first IP address The data is directed to the first acceleration node.

Then, after receiving the ACL policy instruction sent by the network management device, the first network device filters the received data according to the ACL policy instruction. When the first network device receives the data request whose destination address is the first IP address, the first network device directs the data request whose destination address is the first IP address to the first acceleration node. The first acceleration node receives the data request guided by the first network management device through the ACL policy.

This embodiment is applicable to a scenario where a terminal accesses a network through a first network device (eg, MEC or OLT) (eg, a scenario where a home bandwidth accesses the network). The first network device not only serves as a network access device of the first terminal, the first acceleration node deployed in the first network device also serves as a source acceleration node in the overlay network, and the overlay network access points of the terminal are abundant.

In a third implementation manner, the first acceleration node is deployed on a second network device, the second network device is a host in a local area network, and the first terminal is a terminal device in the local area network. The second network device provides the private network AIP for the first acceleration node. In the local area network, the first acceleration node receives the data request from the terminal device through the local area network.

In this embodiment, the acceleration node is embedded and deployed in a local area network (such as an enterprise intranet), and the second network device provides a private network AIP for the first acceleration node, so that terminal devices in the local area network can access the overlay network through the acceleration node . The overlay network provides network acceleration services for terminal devices in the local area network, which solves the problem of expensive public network AIP resources allocated by operators.

In addition, compared with the traditional GA solution, the above three implementation manners do not lose the destination address (eg EIP1) of the original data when the first terminal sends a data request to the first acceleration node. The terminal can access the overlay network only through the public network IP of one acceleration node. One public network IP address can provide shared access for multiple service applications, reducing deployment costs.

Step 508: The first acceleration node obtains a target path, where the target path comes from a routing table entry generated by the central controller.

When the destination address of the data request is the first IP address, the first acceleration node queries the location routing table to determine the second acceleration node corresponding to the first IP address, and the second acceleration node is the destination acceleration node.

For example, the location routing table includes the correspondence between the first IP address and the second acceleration node, as shown in Table 2 below.

Table 2

第一IP地址first IP address	第二加速节点second acceleration node
EIP1EIP1	加速节点DAccelerate Node D
EIP2EIP2	加速节点CAccelerate Node C
IP1IP1	加速节点EAccelerate Node E
IP2IP2	加速节点FAccelerate Node F
……	……

Exemplarily, for the first implementation in the above step 507, the first acceleration node receives the SDK package data, and the first acceleration node decapsulates the SDK package data, and obtains the real purpose IP (such as EIP1 of the original data (data request)) ). The central controller searches the location routing table to determine the acceleration node D that has a corresponding relationship with EIP1.

The first acceleration node sends a data request to the next-hop acceleration node according to the source routing table until the data request is forwarded to the second acceleration node. The source routing table includes the optimal path from the first acceleration node to the second acceleration node. The node is configured to forward the data request to the destination end corresponding to the first IP address.

The first acceleration node queries the source routing table to determine a target path from the first acceleration node to the second acceleration node. For example, the first acceleration node is acceleration node A, the second acceleration node is acceleration node D, and the target path is represented by a list of acceleration nodes (eg acceleration node A, acceleration node B, acceleration node D).

Step 509: The first acceleration node sends a data request to the next-hop acceleration node according to the target path, until the data request is forwarded to the second acceleration node, which is used to forward the data request to the destination.

The first acceleration node performs overlay encapsulation on the original data to obtain overlay encapsulated data. The first acceleration node sends the overlay encapsulation data to the next hop through the overlay tunnel. The overlay package data includes the original data (data request), the target path, the destination address of the next-hop acceleration node, and the source address. The overlay encapsulated data is forwarded hop by hop to the acceleration node on the target path until it is forwarded to the second acceleration node. After the second acceleration node decapsulates the overlay encapsulated data, the data request is sent to the destination.

In the embodiment of the present application, a large number of acceleration nodes are flexibly deployed in the overlay network, so that terminals around the world can access an acceleration node nearby. After the first acceleration node obtains the data request from the terminal, the second acceleration node can be determined by querying the location routing table, and the first acceleration node sends the data request from the terminal to the next-hop acceleration node according to the optimal path indicated in the source routing table. , until the data request is transmitted to the second acceleration node, and the second acceleration node is transmitting the data request to the destination end, so that it is possible to truly realize that terminals worldwide can enjoy the network acceleration service. And compared with the GA scheme in the traditional method, any acceleration node among all the acceleration nodes can be used as a terminal to access the access node of the overlay network, and each acceleration node can be used as a transmission node in the optimal path. The services provided by each acceleration node are shared by all destinations, and there is no need for GA in the traditional method. A business application needs to spend one AIP address mapping, and the project deployment is easy to implement.

Optionally, in order to be able to provide better network forwarding services for the business application, the business application can customize the network setting parameters. The overlay network in this embodiment can provide network acceleration services according to the actual requirements of business applications.

Please refer to FIG. 8 , which is a schematic diagram of two application manners for a business application to apply for a network acceleration service.

The ADN in this application provides network acceleration services for various business applications, and the business applications (such as video service provider A) apply to the central controller for network acceleration services in the following two ways.

Mode 1. The PC of the business application personnel logs in to the console platform. The PC computer responds to the operation of the business application personnel. The business application personnel click on the console interface to select the network setting parameters. The network setting parameters include but are not limited to including acceleration period, bandwidth, cost at least one of. The ADN provides overlay network acceleration services for service applications (such as video service provider A) according to network setting parameters. In this embodiment, the staff of the business application only needs to select according to the network setting parameters provided by the ADN, and the method of applying for the network acceleration service is simple and easy to operate.

Mode 2. The business application directly calls the Tianlu API to apply for the network acceleration service.

The business application and the ADN are in a cooperative relationship, and the AND authorizes the business application, and the business application can directly call the northbound application programming interface (API) of the central controller to customize network parameters. Such as applying for "Optimal QoS", applying for "Optimal Cost", applying for "Optimal QoS + Cost Comprehensive", etc. In this embodiment, the business application can customize the network acceleration service completely according to its own requirements, so as to meet the personalized service requirements of different business applications.

With regard to the above Mode 1, the process of applying for a network acceleration service by a business application is exemplarily described. Please refer to FIG. 9 , which is a schematic diagram of a scenario of an application interface for a network acceleration service. In this application, ADN provides an interface of "application for network acceleration service" for business applications, so that various business applications can apply for network acceleration service. For example, the interface of "Applying for Network Acceleration Service" mainly includes the interface for creating a tenant, creating an acceleration instance (inputting the acceleration instance parameter configuration) interface, adding an acceleration region interface, setting the cloud region domain name interface, and setting the acceleration public network IP address interface, etc. Among them, "setting cloud region domain name interface" is applicable to the scenario where the destination terminal is the cloud region, that is, the scenario where the terminal accesses the cloud region. The "Setting the Acceleration Public Network IP Address Interface" is applicable to the scenario where the destination end is a terminal (or server), that is, the scenario where the terminal accesses the terminal. The steps for applying for the network acceleration service for business applications are as follows: Step a to Step e.

In step a, the PC of the business application personnel displays the console "create user" interface, and the PC of the business application personnel sends user information such as "username" and "password" to the console platform in response to the operation of creating a user by the business application personnel. The console platform sends user information to the central controller.

Step b. The PC of the business application personnel displays the interface of "Create an Acceleration Instance". The "Create Accelerated Instance" interface is used to provide settings for network setting parameters. For example, network setting parameters include bandwidth, acceleration period, and service mode parameters (first mode or second mode). The first mode means that the destination end of the network acceleration service is the cloud region, and the second mode means that the destination end of the network acceleration service is the terminal (or server).

When the first mode is selected, go to the following step c; when the second mode is selected, go to the following step e.

Step c. When the service application personnel select the first mode, the PC of the service application personnel displays the interface of "select acceleration area". The "Select Acceleration Area" interface is used to provide the bound area of the network acceleration service. For example, regions include "Asia", "China", "India", "Europe", etc. The "acceleration area" is used to indicate the area where the user served by the business application is located. For example, if the business application is "NetEase Games", the users of "NetEase Games" are all over the world, and the acceleration area selected by "NetEase Games" may select all regions. While the users of "Video Service Provider A" are mainly distributed in China, the acceleration region selected by "Video Service Provider A" may select all regions as "China". After the terminal responds to the operation of the acceleration area selected by the business application personnel, it sends the target acceleration area (such as China) to the console platform.

In step d, the PC of the business application personnel displays the "cloud region information" interface. The "cloud region information" interface is used to receive the cloud region's identifier, domain name (or EIP). The domain name is the domain name of the service application (eg, video service provider A). After the PC of the business application personnel responds to the operation of the business application personnel, it sends cloud region information to the console platform.

Step e, when the service application personnel select the second mode, the PC of the service application personnel displays an interface of "input acceleration IP". The interface of "Input Acceleration IP" is used to receive the list of public IP addresses of the destination (terminal or server). After the PC responds to the operation of the service application personnel, it sends the public IP of the destination end to the console platform.

After the above steps a to d, or after the steps a, b and e, the console platform establishes an association relationship between the user name and the network setting parameters after receiving the above network setting parameters. The console platform sends the network setting parameters to the central controller through the northbound API.

In the embodiment corresponding to FIG. 5 above, before step 507, the following steps are further included:

The central controller obtains the bandwidth and target acceleration area corresponding to the business application;

The central controller generates committed access rate (CAR) rate limit configuration information according to the bandwidth parameters;

The central controller assigns access rights to the acceleration nodes in the target area according to the EIP information of the destination end, and delivers the CAR speed limit configuration information to the acceleration nodes. The CAR speed limit configuration information is used to guide all acceleration nodes on the optimal path to perform Data forwarding to meet the network acceleration requirements of business applications.

In an application scenario, the application scenario mainly describes a scenario in which the destination end accessed by the first terminal is a cloud region. Referring to Figure 10, after all acceleration nodes are deployed, the central controller issues QoS measurement tasks to all acceleration nodes. Each acceleration node performs QoS measurement on the link status between the acceleration node and neighboring acceleration nodes, and the acceleration node collects the link status information and sends the link status information to the central controller. The central controller calculates the optimal path according to the link state information and the topological structures of all acceleration nodes, and generates a source routing table according to the optimal path. The central controller sends the corresponding source routing table of the acceleration node to each acceleration node.

A business application (eg, video service provider A) applies to the console for a network acceleration service, the destination of the network acceleration service is the cloud region, and the IP information of the cloud region (eg, cloud region1) is EIP1. Optionally, the central controller also acquires information such as the target bandwidth and the target acceleration area corresponding to the "video service provider A".

The central controller determines acceleration node D according to EIP1, and acceleration node D is an acceleration node deployed in cloud region1.

The central controller generates a location routing table, and the location routing table includes the corresponding relationship between the acceleration node D and the EIP1. The central controller sends the location routing table to all acceleration nodes.

The central controller indexes all acceleration nodes in the China region according to the target acceleration region (China region). The central controller generates configuration information according to the target bandwidth.

The central controller sends control information and configuration information to the acceleration nodes in the target acceleration area. The control information includes the EIP of the cloud region (eg, EIP1 ), and the control information is used to instruct the acceleration node to filter the EIP of the cloud region, allowing the data traffic of EIP1 to perform data forwarding according to the configuration information.

The first terminal is connected to the acceleration node A nearby, and the acceleration node A is an acceleration node in the Chinese region. The destination address of acceleration node A's data request is EIP1. The acceleration node A filters data requests whose destination address is EIP1 according to the control information to pass at the rate indicated by the configuration information.

The acceleration node A determines the acceleration node D (ie, the second acceleration node) corresponding to the EIP1 according to the location routing table. The acceleration node A queries the source routing table, and according to the optimal path between the acceleration node A and the acceleration node D in the source routing table (for example, the acceleration node A-acceleration node B-acceleration node C-acceleration node D), the data from the first terminal The data request is forwarded to the next-hop acceleration node B, and the acceleration nodes on the optimal path (such as acceleration node A, acceleration node B, acceleration node C, and acceleration node D) will be sent from the first node according to the configuration information issued by the central controller. The data request of a terminal is forwarded one by one on the optimal path until it is forwarded to the acceleration node D. The acceleration node D forwards the data from the first terminal to the cloud region (EIP1).

In addition, optionally, the network between the POP and the cloud region can implement HBN dedicated line network transmission, or common internet transmission. For example, acceleration node A is deployed in edge cloud A, acceleration node B is deployed in edge cloud B, acceleration node C is deployed at the POP point, and acceleration node D is deployed in cloud region. Then, when the acceleration node C performs overlay encapsulation on the data request, the destination address is the acceleration public network IP address of the acceleration node D, the source address is the acceleration public network IP address of the acceleration node C, and the acceleration node C and the acceleration node D pass through the HBN. The dedicated line network forwards data to improve the network transmission rate from POP to cloud regions. Data can also be transmitted between the acceleration node C and the acceleration node D through the common internet, thereby saving costs for business applications.

In addition, in this application, when the destination terminal is a cloud region, each business application only needs to invoke the network acceleration service provided by AND to realize the terminal's quick access to the cloud region, avoiding repeated and independent development of each business application system.

In the second application scenario, the second application scenario mainly describes a scenario where the destination is a terminal (or server), that is, a scenario of lateral access between terminals. After all acceleration nodes are deployed, the central controller issues QoS measurement tasks to all acceleration nodes. Each acceleration node performs QoS measurement on the link status between the acceleration node and neighboring acceleration nodes, and the acceleration node collects the link status information and sends the link status information to the central controller. The central controller calculates the optimal path in all the acceleration nodes according to the link state information and the topology structure of all the acceleration nodes, and generates a source routing table according to the optimal path. The central controller sends the source corresponding to the acceleration node to each acceleration node. routing table.

When the user applies for the network acceleration service, the console platform obtains the list of public network IP addresses, and sends the list of public network IP addresses to the central controller.

After obtaining the list of public network IP addresses, the central controller queries the IP address database to determine the physical location of each destination (such as the second terminal), and determines the acceleration node closest to the physical location according to the physical location of the destination. For example, IP1 is located in Beijing, and the central controller searches the IP address database to determine that the acceleration node closest to IP1 is acceleration node D (located in Beijing). IP2 is located in Xi'an, and the central controller searches the IP address database to determine that the acceleration node closest to IP2 is acceleration node F (located in Xi'an). The central controller generates a location routing table, and the location routing table includes the correspondence between the public network IP and the acceleration node (eg, the correspondence between IP1 and acceleration node D, and the correspondence between IP2 and acceleration node F).

The central controller sends the location routing table to the acceleration node.

The first terminal accesses the acceleration node A nearby, and the acceleration node A obtains a data request from the first terminal, and the destination IP of the data request is IP1.

The acceleration node A determines, according to the location routing table, that the acceleration node corresponding to IP1 is the acceleration node D (ie, the second acceleration node).

Acceleration node A queries the source routing table, and forwards the data request to the optimal path between acceleration node A and acceleration node D in the source routing table (such as acceleration node A-acceleration node B-acceleration node C-acceleration node D) The next hop accelerates the node B, and the data request is forwarded hop by hop on the optimal route until it is forwarded to the acceleration node D. The acceleration node D forwards the data from the first terminal to the second terminal.

Compared with the GA in the traditional method, the GA in the traditional method only supports the scenario in which the terminal accesses the cloud region, while the ADN in this embodiment not only supports the scenario in which the terminal accesses the cloud region, but also supports the scenario in which the terminal and the terminal are accessed. Universality.

Exemplarily, with respect to step 509 in the embodiment corresponding to FIG. 5 , the process of forwarding the data request from the terminal to the destination terminal by the first acceleration node will be described. 1. The destination is the cloud region, which is the first application scenario above. 2. The destination end is a terminal (or server), that is, the second application scenario above.

First, the data encapsulation format will be described. Please refer to FIG. 11 , which is a schematic diagram of a data format for overlay encapsulation based on a user datagram protocol (UDP). The packets transmitted in the overlay tunnel encapsulate the original data to obtain overlay encapsulation data. The format of the overlay encapsulation data includes the following fields.

IP header field: includes the source address (32 bits in length) and the destination address (32 bits in length).

The UDP header field includes the source port number (16 bits in length) and the destination port number (16 bits in length).

Segment list (segment list, SL) field: used to indicate the nodes that the data packet needs to pass through during the forwarding process. The list is segment list[0] to segment list[n-1]. Among them, [*] is used to represent the node number (or also called "subscript"), and n represents the number of accelerated nodes in the optimal path. When the source acceleration node presses the path label, it will press the multi-layer label, that is, the node that needs to be passed during the transmission of the data packet. For example, the optimal path includes n nodes (such as node A, node B, node C, etc.), and the first one pushed into the destination address is segment list[n-1] (for example, segment list[2]) corresponds to The IP address of the acceleration node. The last one pushed into the destination address is the IP address of segment list[0]. For example, the segment list can look like this.

segment list[0]=IPD;

segment list[1]=IPC;

segment list[2]=IPB.

The first segment field: 8 bits in length, used to refer to the first hop through which data is sent from the source acceleration node to the destination acceleration node. In the segment list, the bottom node (segment list[n-1]) is the node closer to the source acceleration node, and the top one is the destination acceleration node segment list[0]), so the value of the first segment field is " n-1".

The remaining segment (segment left) field: used to indicate the currently active segment, that is, used to indicate the next hop where data will be transmitted. When the source acceleration node obtains the data, the data has not yet been transmitted to the next hop, and the value of the segment left field is n-1, indicating that the next node to pass through is the node corresponding to segment list[n-1] (such as the node B).

Each time the data passes through a node in the segment list, the field value of segment left will decrease by 1. The acceleration node will copy the IP address of the node of the segment list [SL] to the destination address field in the packet header, thereby indicating the next hop node and sending the data to the destination. For example, A is the source acceleration node, D is the destination acceleration node, and the acceleration nodes that the optimal path passes through are B, C, and D. At this time, the acceleration node B is the first acceleration node to pass from the acceleration node A to the acceleration node D, so the value of the first segment field is the subscript "2" corresponding to the acceleration node B. At this time, the data has not been sent, that is, the nodes in the segment list have not yet arrived, so the value of the segment left field is also the subscript "2" corresponding to the acceleration node B. At the same time, the source acceleration node will copy the address (IPB) of Segment List[2]=B to the destination address of the packet header. When the data packet arrives at the acceleration node B, the acceleration node B checks that the destination address of the header is IPB after receiving the data. The acceleration node B removes the header, and the value of the segment left field is obtained as "2". The acceleration node B determines that the transmitted data has not reached the destination acceleration node. The acceleration node B also needs to continue to forward the data, and the acceleration node B keeps the first segment. The value of the field does not change (eg 2). The acceleration node B decrements the value of segment left by 1, and changes the value of the segment left field to 1. And, the acceleration node B copies the address corresponding to the segment list[SL=1] (the IP address of the acceleration node C) to the destination address, and then continues to forward the data.

Payload length field: The length is 16 bits.

In this embodiment, the overlay tunnel encapsulates IP Layer 3 packets in UDP mode for data forwarding, and the data in the IP packet (that is, the original data) may be various transmissions such as transmission control protocol (TCP) or UDP. Type of data packets, overlay encapsulated data packets can not be constrained by transmission type and application type, and the network acceleration service has a wider scope.

Next, the process of data transfer will be described. The process of data forwarding is divided into the case where the destination is a cloud region, and the case where the destination is a terminal (or server).

1. The destination is the cloud region, that is, the scenario where the first terminal accesses the cloud region. Exemplarily, the access by the first terminal to the first acceleration node is taken through the SDK tunnel as an example, the first acceleration node is the acceleration node A as an example, the IP address of the first terminal is IP1, and the public network of the acceleration node A is used as an example. An IP address is an IPA. The central controller pre-configures the first NAT IP (also called "first NAT IP") for the source acceleration node and the destination acceleration node, and configures the second NAT IP (also called "tail NAT IP") for the destination acceleration node. For example, the first NAT IP is IP8 and the second NAT IP is IP9. Exemplarily, in the overlay network, the optimal path from the source acceleration node to the destination acceleration node is: acceleration node A → acceleration node B → acceleration node D. The acceleration node A is configured with a public network IPA and a first NAT IP (IP8), and IPA and IP8 can be different IP addresses, or, in order to save public network IP, IPA and IP8 can be the same IP address. The acceleration node D is configured with the public network IPD and the first NAT IP (IP9). The IPD and the IP9 can be different IP addresses, or, in order to save the public network IP, the IPD and the IP9 can be the same IP address.

Exemplarily, please refer to FIG. 12 , which is a schematic diagram of a scenario of data forwarding between the first terminal and the cloud region.

S41. The first terminal sends a data packet to the acceleration node A, the destination address of the data packet is EIP, and the source address is IP1. The first terminal performs SDK encapsulation on the data packet to obtain SDK encapsulation data. The destination address of the SDK encapsulation data is IPA, and the source address of the SDK encapsulation data is IP1. In the first case, when IP1 is a public network IP, the first terminal sends the SDK package data to acceleration node A. After acceleration node A receives the SDK package data, it strips off the SDK header and exposes the destination address of the original data as EIP, the source address is IP1. In the second case, when IP1 is a private network IP address, the SDK encapsulates the data through the network address translation (NAT) device of the operator's network to reach acceleration node A. The public IP address of the NAT device is IPM.

S42. The acceleration node A strips off the SDK header to expose the original data, and the acceleration node A modifies the source address in the inner layer to the public network IPM after network address translation.

S43. The acceleration node A performs source address translation (source NAT, SNAT), and converts the IPM into the first NAT IP (such as IP8). That is, the source address in the inner layer is IP8 and the destination address is EIP. It can be understood that in this step, SNAT is the port mapping, and the IPM is mapped to IP8. In this step, the purpose of the acceleration node A converting the IPM to the first NAT IP is to use IP8 as the IP address of the destination acceleration node when the cloud region returns the data stream.

S44. The acceleration node A performs overlay encapsulation on the inner layer data. The packet format of the overlay encapsulation data is shown in Figure 11. The overlay encapsulation data includes the inner layer data (that is, the original data, the source address is IP8, and the destination address is EIP), The IP address (IPB) of the next hop in the overlay header and the IP address of the acceleration node in the optimal path (for example, the IP address of acceleration node A is IPA, the IP address of acceleration node B is IPB, and the IP address of acceleration node D is IPD).

S45. After the acceleration node B receives the overlay encapsulation data, it determines that the data packet does not reach the destination acceleration node according to the value of the segment left field, and the acceleration node B continues to modify the IP address of the next hop in the overlay encapsulation data to IPD, and changes the The overlay encapsulated data is forwarded to the next hop (acceleration node D).

S46. After the acceleration node D receives the overlay package data, it determines that the overlay package data has reached the destination acceleration node according to the value of the segment left field. After the acceleration node D strips the header of the overlay package data, the source address of the exposed inner layer data is IP8, the destination address is EIP. The acceleration node D maps the source address of the inner data to the tail NAT IP (such as IP9) after going through SNAT. The acceleration node D is the acceleration node deployed in the cloud region (destination terminal), and the tail NAT IP is the IP address assigned by the cloud region. Optionally, in order to save the public network IP address, the IPD of the acceleration node D and the tail NAT IP can be the same IP. The acceleration node D accesses the data center through the cloud region internal network (IP information is EIP). In this step, the purpose of the acceleration node D mapping the destination address of the inner layer data to the tail NAT IP (such as IP9) is to use IP9 as the IP address of the source acceleration node when the cloud region returns the data stream.

The above steps S41-S46 are the forward data traffic forwarding process, that is, the process in which the first terminal sends data to the cloud region. The following steps S51-S56 are the reverse data traffic forwarding process, that is, the process in which the cloud region sends data to the first terminal.

S51. The data center sends the feedback original data to the acceleration node D. The destination address of the feedback original data is IP9, and the source and destination address is EIP, that is, the data center uses the tail NAT IP (IP9) as the destination address, and the data center connects to the cloud The acceleration node D of the region.

S52. After the acceleration node D receives the feedback original data, it performs destination address translation on the destination address in the feedback original data, and maps the destination address to IP8.

S53. The acceleration node D searches the location routing table, and determines the acceleration node A corresponding to IP8 according to the location routing table. The acceleration node D determines the optimal path (the list of acceleration nodes) from the acceleration node D to the acceleration node A according to the source routing table. Acceleration node D overlay-encapsulates the optimal path (list of acceleration nodes), the next-hop acceleration node (such as acceleration node B), and the feedback original data, and forwards the overlay-encapsulated data hop by hop until it is forwarded to acceleration node A .

S54. After the acceleration node A receives the overlay package data, it strips off the overlay header to expose the destination address (IP8) and source address (EIP) of the inner layer feedback data.

S55. The acceleration node A performs destination address translation, and maps IP8 to the public network IPM of the NAT device.

S56. The NAT device maps the IPM to the private network IP address (IP1) of the first terminal, and the NAT device forwards the private network.

2. The destination end is a terminal (or server), that is, a scenario in which the first terminal accesses the second terminal (or server).

Please refer to FIG. 13 . FIG. 13 is a schematic diagram of a scene diagram of data forwarding between the first terminal and the second terminal. The first terminal has a built-in SDK plug-in, and the first terminal can access the central controller through the SDK plug-in. The second terminal has a built-in SDK plug-in, and the second terminal can access the central controller through the SDK plug-in. Exemplarily, the IP address of the first terminal is IP2, and the IP address of the second terminal is IP3.

The first terminal accesses the central controller through the SDK, and the central controller feeds back the IP address of the acceleration node A to the first terminal. The first terminal accesses the acceleration node A (source acceleration node).

S61. The first terminal obtains original data, the destination address of the original data is IP3, and the source address is IP2.

S62. The first terminal performs SDK encapsulation on the original data to be sent, and sends the SDK encapsulated data to acceleration node A through the SDK tunnel. SDK package data includes SDK header and original data. The source address in the SDK header is IP2, and the destination address is IPA (the public IP of the acceleration node A).

S63. The acceleration node A decapsulates the SDK package data, strips off the SDK header, and exposes the destination address (IP3) and source address (IP2) of the original data. The acceleration node A searches the location routing table, and determines the acceleration node (eg, the acceleration node D) that has a corresponding relationship with the destination address (IP3). The acceleration node A searches the source routing table to determine the optimal path from the acceleration node A to the acceleration node D, that is, the acceleration node (segment list) that needs to be experienced from the acceleration node A to the acceleration node D. Acceleration node A performs overlay encapsulation on the original data packet to obtain overlay encapsulation data. The overlay encapsulation data includes the original data, the IP of each acceleration node on the optimal path (for example, IPA, IPB, IPC, and IPD), and the next-hop acceleration node. IP (eg IPB). The acceleration node A sends the overlay encapsulated data, and the overlay encapsulated data is forwarded hop by hop until forwarded to the acceleration node D (destination forwarding node).

S64. The acceleration node D decapsulates the overlay package data to obtain original data.

S65. The acceleration node D performs SDK encapsulation on the original data to obtain SDK encapsulated data. The outer destination address of the SDK encapsulated data is IP3, and the source address is IPD. The acceleration node D sends the SDK package data to the second terminal through the SDK tunnel.

S66. The second terminal decapsulates the SDK package data to obtain original data.

The above steps S61-S66 are exemplary descriptions of the forward forwarding process of the data flow, that is, the process of sending data from the first terminal to the second terminal. The following steps S71-S76 are the reverse data traffic forwarding process, that is, the process in which the second terminal sends data to the first terminal. The second terminal accesses the central controller through the SDK, and the central controller feeds back the IP address of the acceleration node D to the second terminal. The first terminal accesses the acceleration node D (source acceleration node).

S71. The second terminal obtains the raw data fed back, the destination address of the raw data fed back is IP2, and the source address is IP3.

S72. The second terminal performs SDK encapsulation on the feedback original data, and sends the SDK encapsulation data to acceleration node D through the SDK tunnel. SDK package data includes SDK header and feedback data. The source address in the header of the SDK package data is IP3, and the destination address is IPD (the public IP of the acceleration node D).

S73. The acceleration node D decapsulates the SDK encapsulated data, strips off the header, and exposes the destination address (IP2) and source address (IP3) of the fed back original data. The acceleration node D searches the location routing table, and determines the acceleration node (eg, the acceleration node A) that has a corresponding relationship with the destination address (IP2). The acceleration node D searches the source routing table to determine the optimal path from the acceleration node D to the acceleration node A, that is, the acceleration node (segment list) that needs to be experienced from the acceleration node D to the acceleration node A. The acceleration node D performs overlay encapsulation on the feedback data to obtain the overlay encapsulation data. The overlay encapsulation data includes the feedback data, the IP of each acceleration node (for example, IPD, IPC, IPB, and IPA) on the optimal path, and the IP address of the next-hop acceleration node. IP (eg IPC). The acceleration node D sends the overlay encapsulated data, and the overlay encapsulated data is forwarded hop by hop until forwarded to the acceleration node A (destination forwarding node).

S74. The acceleration node A decapsulates the overlay package data, and obtains the original data fed back.

S75. The acceleration node A performs SDK encapsulation on the feedback data. The outer destination address of the SDK encapsulated data is IP2 and the source address is IPA. The acceleration node A sends the SDK package data to the first terminal through the SDK tunnel.

S76. The first terminal decapsulates the SDK package data, and obtains the original data that is fed back.

Optionally, in order to ensure the security of the forwarded data, the data transmitted between the acceleration nodes is encrypted and encapsulated, and the transmission is encrypted to prevent the data from being eavesdropped. Please refer to FIG. 14 . FIG. 14 is a schematic diagram of encrypted and forwarded data in an overlay tunnel.

First, the encryption package. When the source acceleration node (such as acceleration node A) obtains the original data, the encryption key field is added, and the original data is encrypted and filled to obtain encrypted data.

Then, encrypt the transmission. The source acceleration node sends encrypted data to the next-hop acceleration node, the encrypted data is forwarded hop by hop, and the encrypted data is kept encrypted during the forwarding process. Until the encrypted data reaches the destination acceleration node (such as acceleration node D).

The acceleration node D performs overlay decapsulation on the encrypted data, and also de-encrypts the data, restores the original data, and forwards the original data.

The GA in the traditional method relies on the IP private line network of the cloud manufacturer to provide acceleration services, and can only encrypt the data to be transmitted through the application layer, while the transport layer does not support encryption services. In this application, AND forwards data based on the overlay tunnel, and can naturally encrypt data based on the overlay tunnel, so that some confidential data can be protected by double-layer encryption at the application layer and the transport layer, thereby ensuring data security.

Exemplarily, the acceleration effect of the data forwarding method in the embodiment of the present application is exemplarily described. Please refer to Table 3 below, which is the test data of the data forwarding acceleration effect.

table 3

It should be understood that the data in the above Table 3 are only illustrative, not all test data. As can be seen from Table 3 above, the acceleration effect of different networking acceleration nodes in AND varies, and the better optimization effect can reach more than 40%. Most of the optimal paths only need to go through one acceleration node to achieve delay. optimal. In the embodiment of the present application, the network acceleration service does not depend on the dedicated line network, and for the Internet network that is universally covered in the world, AND can achieve a better acceleration optimization effect.

As shown in FIG. 2, the communication system includes a central controller and a plurality of acceleration nodes, and the plurality of acceleration nodes include a first acceleration node and a second acceleration node, wherein the deployment environment of the central controller belongs to the first acceleration node. A cloud service provider, the deployment environment of the multiple acceleration nodes belongs to a second cloud service provider, an application service provider or a telecom operator. An embodiment of the present application provides an acceleration node. The acceleration node 1500 is described by taking a first acceleration node as an example, and the first acceleration node may be any acceleration node among multiple acceleration nodes. The first acceleration node is configured to implement the functions performed by the first acceleration node in the foregoing method embodiments. Referring to FIG. 15 , the acceleration node 1500 includes a forwarding module 1501 and a control module 1502, wherein the forwarding module 1501 is used to implement the function of the forwarding node in the above method embodiment, and the control module 1502 is used to implement the local control in the above method embodiment. function of the device.

A forwarding module 1501, configured to receive a data request from the first terminal, where the data request is used to access the destination;

a control module 1502, configured to obtain a target path, the target path is from a routing table entry generated by the central controller;

The forwarding module 1501 is configured to send the data request to the next-hop acceleration node according to the target path, until the data request is forwarded to the second acceleration node, and the second acceleration node is configured to forward the data The request is forwarded to the destination.

Further, the forwarding module 1501 is configured to perform step 507 and step 509 in the above-mentioned embodiment corresponding to FIG. 5 . When the first acceleration node is the acceleration node A, the forwarding module 1501 is further configured to perform steps S42 , S43 and S44 in the example corresponding to FIG. 12 . When the first acceleration node is the acceleration node B, the forwarding module 1501 is further configured to perform step S45 in the example corresponding to FIG. 12 . When the first acceleration node is the acceleration node D, the forwarding module 1501 is further configured to perform steps S46 , S51 , S52 and S54 in the example corresponding to FIG. 12 . When the first acceleration node is the acceleration node A, the forwarding module 1501 is further configured to execute steps S63, S74 and S75 in the example corresponding to FIG. 13 , and when the first acceleration node is the acceleration node D, the forwarding module 1501 is further configured to execute Steps S64, S65 and S73 in the example corresponding to FIG. 13 . The control module 1502 is configured to execute step 508 in the embodiment corresponding to FIG. 5 , and steps S53 , S55 and S56 in the example corresponding to FIG. 12 .

Specifically, in an optional implementation manner, the routing table entry includes a source routing table and a location routing table; the control module 1502 is further configured to receive the source routing table and the location routing table sent by the central controller, where the source routing table includes The path from the source acceleration node to the destination acceleration node. The location routing table includes the correspondence between the first IP address and the second acceleration node. The first IP address is the IP address of the destination end where the user applies for the network acceleration service; when the destination address of the data request When it is the first IP address, the first acceleration node queries the location routing table to determine the second acceleration node corresponding to the first IP address, and the second acceleration node is the destination acceleration node; when the first acceleration node is the source acceleration node, the first acceleration node is the source acceleration node. An acceleration node queries the source routing table according to the second acceleration node, and determines a target path from the first acceleration node to the second acceleration node.

In an optional implementation manner, the forwarding module 1501 is further configured to measure the link status between the first acceleration node and the neighboring acceleration node to obtain link status information; the control module 1502 is further configured to send the information to the central controller The link state information obtained by the forwarding module 1501 is used for the central controller to generate the source routing table.

In an optional implementation manner, an SDK plug-in is configured in the first terminal, and the address of the central controller is preset in the SDK plug-in; the forwarding module 1501 is further configured to receive SDK package data from the first terminal through the SDK tunnel, The SDK encapsulation data is the data after encapsulating the data request. The destination address in the header of the SDK encapsulation data is the IP address of the first acceleration node, and the source address in the header is the IP address of the first terminal.

In an optional implementation manner, the deployment environment of the first acceleration node is a first network device, the first network device is used to receive an access control list ACL policy instruction, and the ACL policy instruction is used to trigger the first network device to convert the destination address The data of the first IP address is directed to the first acceleration node; the forwarding module 1501 is further configured to receive a data request from the first terminal directed by the first network device according to the ACL policy instruction.

In an optional implementation manner, the deployment environment of the first acceleration node is a device in a local area network, and the first terminal is a terminal in the local area network; the forwarding module 1501 is further configured to receive a data request from the first terminal through the local area network.

In one possible design, the acceleration node 1500 runs in a virtual machine or container provided by the deployment environment. Please refer to FIG. 16 , which is a schematic diagram of the architecture of a virtual machine. The architecture of the virtual machine includes a hardware layer 1601 , a virtualization layer 1602 and a virtual machine 1603 . The virtualization layer 1602 includes a hypervisor. The hypervisor is used to manage the real hardware resources of the hardware layer 1601 , and provides hardware resource abstraction for the virtual machine 1603 , thereby providing a running environment for the acceleration node 1500 in the virtual machine 1603 . Hardware layer 1601 may include one or more processors, memory, and storage devices. The storage device and the memory are both connected to the processor. The processor can also be referred to as a processing unit, which can implement certain control functions. The processor may be a general-purpose processor or a special-purpose processor, or the like. Instructions may be stored on the memory, and the instructions may be executed on the processor. The storage device is used to store the source routing table and the location routing table. The hypervisor provides hardware resource abstraction for the virtual machine, so that the acceleration node in the virtual machine executes the method executed by the first acceleration node in the above method embodiment.

Referring to FIG. 17 , an embodiment of the present application further provides a central controller, where the central controller is configured to execute the method executed by the central controller in the foregoing method embodiments. The central controller 1700 includes a transceiver module 1701 and a processing module 1702 .

A transceiver module 1701, configured to acquire link status information sent by the acceleration node;

The processing module 1702 is configured to generate a source routing table according to the link state information obtained by the transceiver module 1701, where the source routing table includes a path from the source acceleration node to the destination acceleration node;

The transceiver module 1701 is used to obtain the first IP address of the destination end of the user applying for the network acceleration service;

The processing module 1702 is further configured to generate a location routing table, where the location routing table includes the correspondence between the first IP address and the second acceleration node;

The transceiver module 1701 is further configured to send the location routing table and the source routing table corresponding to the first acceleration node to the first acceleration node. The location routing table is used to guide the first acceleration node to determine the second acceleration node according to the first IP address. The IP address is the destination address of the data request from the first terminal, the source routing table is used by the first acceleration node to obtain the target path, and the target path is used to guide the data request to be forwarded to the second acceleration node, and the second acceleration node is used to forward the data to the destination.

Further, optionally, the transceiver module 1701 is a transceiver. Among them, the transceiver has the function of sending and/or receiving. Optionally, the transceiver is replaced by a receiver and/or a transmitter.

Optionally, the transceiver module 1701 is a communication interface. Optionally, the communication interface is an input-output interface or a transceiver circuit. The input and output interface includes an input interface and an output interface. The transceiver circuit includes an input interface circuit and an output interface circuit.

Optionally, the processing module 1702 is a processor, and the processor is a general-purpose processor or a special-purpose processor or the like. Optionally, the processor includes a transceiver unit for implementing receiving and transmitting functions. For example, the transceiver unit is a transceiver circuit, or an interface, or an interface circuit. Transceiver circuits, interfaces, or interface circuits for implementing receiving and transmitting functions are deployed separately, or optionally, integrated together. The above-mentioned transceiver circuit, interface or interface circuit is used for reading and writing code or data, or the above-mentioned transceiver circuit, interface or interface circuit is used for signal transmission or transmission.

Further, the transceiver module 1701 is configured to execute step 501 , step 503 , step 504 and step 506 in the above-mentioned embodiment corresponding to FIG. 5 . The processing module 1702 is configured to execute step 502 and step 505 in the above-mentioned embodiment corresponding to FIG. 5 .

Specifically, in an optional implementation manner, the processing module 1702 is further specifically configured to: determine the second acceleration node according to the first IP address of the destination; establish a correspondence between the first IP address and the second acceleration node; Relationships generate location routing tables.

In an optional implementation manner, when the destination terminal is the cloud area, the processing module 1702 is further configured to determine the second acceleration node deployed in the cloud area according to the first IP address.

In an optional implementation manner, when the destination terminal is a terminal or a server, the processing module 1702 is further configured to query the IP address database according to the first IP address to determine the physical location of the destination terminal; determine the second closest to the physical location. Speed up nodes.

In an optional implementation manner, an SDK plug-in is configured in the first terminal, and the address information of the central controller is preset in the SDK plug-in; the transceiver module 1701 is further configured to receive a request sent by the first terminal; The first terminal feeds back the IP address of the first acceleration node, and the IP address of the first acceleration node is used by the first terminal to send a data request to the first acceleration node by using the SDK tunnel.

In an optional implementation manner, the transceiver module 1701 is further configured to send traffic diversion information to the network management system, the traffic diversion information includes IP information of the destination end, and the traffic diversion information is used to trigger the network management system to send an ACL policy instruction to the first network device , the second acceleration node is an acceleration node deployed in the first network device, and the ACL policy instruction is used to trigger the first network device to direct the data request from the first terminal to the first acceleration node.

In an optional implementation manner, the transceiver module 1701 is further configured to acquire a mode parameter, where the mode parameter includes a first mode and a second mode, wherein the first mode is used to indicate that the destination of the network acceleration service is a cloud area, The second mode is used to indicate that the destination of the network acceleration service is the second terminal or the server.

Referring to FIG. 18 , an embodiment of the present application provides a central controller, and the central controller 1800 is used to implement the method executed by the central controller in the above method embodiments. For details, please refer to the descriptions in the above method embodiments.

The central controller 1800 may include one or more processors 1801, and the processors 1801 may also be referred to as processing units, which may implement certain control functions. The processor 1801 may be a general-purpose processor or a special-purpose processor, or the like. The central processing unit can be used to control the central controller, execute software programs, and process data of the software programs.

In an optional design, the processor 1801 may also store instructions 1803, and the instructions 1803 may be executed by the processor, so that the central controller 1800 executes the methods described in the above method embodiments.

In another optional design, the processor 1801 may include a transceiver unit for implementing the functions of receiving and transmitting. For example, the transceiver unit may be a transceiver circuit, or an interface, or an interface circuit. Transceiver circuits, interfaces or interface circuits used to implement receiving and transmitting functions may be separate or integrated. The above-mentioned transceiver circuit, interface or interface circuit can be used for reading and writing code/data, or the above-mentioned transceiver circuit, interface or interface circuit can be used for signal transmission or transmission.

In another possible design, the central controller 1800 may include a circuit, and the circuit may implement the function of sending or receiving in the above method embodiments.

The central controller 1800 may include one or more memories 1802 on which instructions 1804 may be stored, and the instructions may be executed on the processor, so that the central controller 1800 executes the methods described in the above method embodiments. Optionally, data may also be stored in the memory. Optionally, instructions and/or data may also be stored in the processor. The processor and the memory can be provided separately or integrated together.

Optionally, the central controller 1800 may further include a transceiver 1805 and/or an antenna 1806 . The processor 1801 may be called a processing unit, and controls the central controller 1800 . The transceiver 1805 may be referred to as a transceiver unit, a transceiver, a transceiver circuit, a transceiver device or a transceiver module, etc., and is used to implement a transceiver function.

An embodiment of the present application is a computer program product, the computer program product includes computer program code, and when the computer program code is executed by a computer, enables the computer to implement the method executed by the central controller in the above method embodiments.

An embodiment of the present application is a computer program product. The computer program product includes computer program code, which, when executed by a computer, enables the computer to implement the method executed by the first acceleration node in the above method embodiments.

An embodiment of the present application is a computer-readable storage medium for storing a computer program or instruction, which, when executed, causes the computer to execute the method executed by the central controller in the above method embodiment.

An embodiment of the present application is a computer-readable storage medium for storing a computer program or instruction, and when the computer program or instruction is executed, the computer executes the method executed by the first acceleration node in the foregoing method embodiment.

As mentioned above, the above embodiments are only used to illustrate the technical solutions of the present application, but not to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand: The technical solutions described in the embodiments are modified, or some technical features thereof are equivalently replaced; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions in the embodiments of the present application.

Claims

A data forwarding method, characterized by being applied to a first acceleration node in a communication system, the communication system comprising a central controller and a plurality of acceleration nodes, the plurality of acceleration nodes including the first acceleration node and the first acceleration node Two acceleration nodes, wherein the deployment environment of the central controller belongs to a first cloud service provider, and the deployment environments of the multiple acceleration nodes belong to a second cloud service provider, an application service provider or a telecom operator;

The first acceleration node receives a data request from the first terminal, where the data request is used to access the destination;

The first acceleration node obtains a target path, and the target path comes from a routing table entry generated by the central controller;

The first acceleration node sends the data request to the next-hop acceleration node according to the target path, until the data request is forwarded to the second acceleration node, and the second acceleration node is used for converting the data The request is forwarded to the destination.
The method according to claim 1, wherein the routing table entry comprises a source routing table and a location routing table; the method further comprises:

The first acceleration node receives the source routing table and the location routing table sent by the central controller, the source routing table includes a path from the source acceleration node to the destination acceleration node, and the location routing table includes the first IP address and the location routing table. The corresponding relationship of the second acceleration node, the first IP address is the IP address of the destination end where the user applies for the network acceleration service;

The first acceleration node obtains the target path, including:

When the destination address of the data request is the first IP address, the first acceleration node queries the location routing table to determine a second acceleration node corresponding to the first IP address, and the second acceleration node The node is the destination acceleration node;

When the first acceleration node is a source acceleration node, the first acceleration node queries the source routing table according to the second acceleration node, and determines a destination from the first acceleration node to the second acceleration node path.
The method according to claim 2, wherein before the first acceleration node receives the source routing table sent by the central controller, the method further comprises:

The first acceleration node measures a link state between the first acceleration node and a neighbor acceleration node to obtain link state information;

The first acceleration node sends the link state information to the central controller, where the link state information is used by the central controller to generate the source routing table.
The method according to any one of claims 1-3, wherein an SDK plug-in is configured in the first terminal, the address of the central controller is preset in the SDK plug-in, and the first terminal is configured with an SDK plug-in. The acceleration node receives the data request from the first terminal, including:

The first acceleration node receives SDK package data from the first terminal through the SDK tunnel, the SDK package data is the data after the data request is packaged, and the destination address in the header of the SDK package data is The IP address of the first acceleration node, and the source address in the header is the IP address of the first terminal.
The method according to any one of claims 1-3, wherein the deployment environment of the first acceleration node is a first network device, and the first network device is configured to receive an access control list (ACL) policy instruction, The ACL policy instruction is used to trigger the first network device to guide the data whose destination address is the first IP address to the first acceleration node; the first acceleration node receives the data request from the first terminal, including:

The first acceleration node receives a data request from the first terminal guided by the first network device according to the ACL policy instruction.
The method according to any one of claims 1-3, wherein the deployment environment of the first acceleration node is a device in a local area network, the first terminal is a terminal in the local area network, and the first acceleration node is a terminal in the local area network. An acceleration node receives the data request from the first terminal, including:

The first acceleration node receives a data request from the first terminal through the local area network.
The method according to claim 1, wherein the destination end is a cloud area, or the destination end is a second terminal or a server.
The method according to any one of claims 1-7, wherein the deployment environment includes a cloud area, an access point POP, an edge cloud, an optical line terminal OLT, or a multi-access edge computing device MEC.
A data forwarding method, characterized in that it is applied to a central controller in a communication system, the communication system comprising a central controller and a plurality of acceleration nodes deployed in various deployment environments, the plurality of acceleration nodes comprising a first An acceleration node and a second acceleration node, the method includes:

obtaining, by the central controller, the link status information sent by the acceleration node;

The central controller generates a source routing table according to the link state information, and the source routing table includes a path from the source acceleration node to the destination acceleration node;

The central controller obtains the first IP address of the destination end where the user applies for the network acceleration service;

The central controller generates a location routing table, and the location routing table includes the correspondence between the first IP address and the second acceleration node;

The central controller sends the location routing table and the source routing table corresponding to the first acceleration node to the first acceleration node, where the location routing table is used to guide the first acceleration node according to the first IP address Determine the second acceleration node, the first IP address is the destination address of the data request from the first terminal, the source routing table is used for the first acceleration node to obtain a target path, and the target path is used for guiding The data request is forwarded to the second acceleration node, and the second acceleration node is configured to forward the data to the destination.
The method of claim 9, wherein the central controller generates a location routing table, comprising:

The central controller determines the second acceleration node according to the first IP address of the destination;

The central controller establishes a correspondence between the first IP address and the second acceleration node;

The central controller generates a location routing table according to the corresponding relationship.
The method according to claim 10, wherein when the destination terminal is a cloud area, the central controller determines the second acceleration node according to the first IP address of the destination terminal, comprising:

The central controller determines the second acceleration node deployed in the cloud area according to the first IP address.
The method according to claim 10, wherein when the destination is a terminal or a server, the central controller determines the second acceleration node according to the first IP address of the destination, comprising:

The central controller queries an IP address database according to the first IP address, and determines the physical location of the destination;

The central controller determines the second acceleration node closest to the physical location.
The method according to any one of claims 10-12, wherein an SDK plug-in is configured in the first terminal, and address information of the central controller is preset in the SDK plug-in; the method Also includes:

receiving, by the central controller, the request sent by the first terminal;

The central controller feeds back the IP address of the first acceleration node to the first terminal according to the request, and the IP address of the first acceleration node is used by the first terminal to use the SDK tunnel to report to the first terminal. The acceleration node sends the data request.
The method according to any one of claims 9-12, wherein the method further comprises:

The central controller sends traffic diversion information to the network management system, where the traffic diversion information includes IP information of the destination terminal, and the traffic diversion information is used to trigger the network management system to send an ACL policy instruction to the first network device, and the first The second acceleration node is an acceleration node deployed in the first network device, and the ACL policy instruction is used to trigger the first network device to direct the data request from the first terminal to the first acceleration node.
The method according to claim 9, wherein the method further comprises:

The central controller acquires a mode parameter, the mode parameter includes a first mode and a second mode, wherein the first mode is used to indicate that the destination of the network acceleration service is a cloud area, and the second mode is used to indicate The destination end of the network acceleration service is the second terminal or server.
An acceleration node, characterized in that it is included in a communication system, the communication system includes a central controller and a plurality of the acceleration nodes, and the plurality of the acceleration nodes includes a first acceleration node and a second acceleration node, wherein the The deployment environment of the central controller belongs to a first cloud service provider, and the deployment environment of the multiple acceleration nodes belongs to a second cloud service provider, an application service provider or a telecom operator; the first acceleration node includes:

a forwarding module, configured to receive a data request from the first terminal, where the data request is used to access the destination;

a control module, configured to obtain a target path, the target path is from a routing table entry generated by the central controller;

The forwarding module is configured to send the data request to the next-hop acceleration node according to the target path until the data request is forwarded to the second acceleration node, and the second acceleration node is configured to forward the data request to the second acceleration node. The data request is forwarded to the destination.
The acceleration node according to claim 16, wherein the routing table entry comprises a source routing table and a location routing table;

The control module is further configured to receive a source routing table and a location routing table sent by the central controller, where the source routing table includes a path from a source acceleration node to a destination acceleration node, and the location routing table includes a first IP address The corresponding relationship between the address and the second acceleration node, the first IP address is the IP address of the destination end where the user applies for the network acceleration service;

When the destination address of the data request is the first IP address, the first acceleration node queries the location routing table to determine a second acceleration node corresponding to the first IP address, and the second acceleration node The node is the destination acceleration node;

When the first acceleration node is a source acceleration node, the first acceleration node queries the source routing table according to the second acceleration node, and determines a destination from the first acceleration node to the second acceleration node path.
The acceleration node according to claim 17, wherein,

The forwarding module is further configured to measure the link state between the first acceleration node and the neighbor acceleration node to obtain link state information;

The control module is further configured to send the link state information obtained by the forwarding module to the central controller, where the link state information is used by the central controller to generate the source routing table.
The acceleration node according to any one of claims 16-18, wherein an SDK plug-in is configured in the first terminal, and an address of the central controller is preset in the SDK plug-in;

The forwarding module is further configured to receive SDK encapsulation data from the first terminal through the SDK tunnel, where the SDK encapsulation data is the data after encapsulating the data request, and the purpose in the header of the SDK encapsulation data The address is the IP address of the first acceleration node, and the source address in the header is the IP address of the first terminal.
The acceleration node according to any one of claims 16-18, wherein a deployment environment of the first acceleration node is a first network device, and the first network device is configured to receive an access control list (ACL) policy instruction , the ACL policy instruction is used to trigger the first network device to guide the data whose destination address is the first IP address to the first acceleration node;

The forwarding module is further configured to receive a data request from the first terminal guided by the first network device according to the ACL policy instruction.
The acceleration node according to any one of claims 16-18, wherein the deployment environment of the first acceleration node is a device in a local area network, and the first terminal is a terminal in the local area network;

The forwarding module is further configured to receive a data request from the first terminal through the local area network.
A central controller, comprising:

The transceiver module is used to obtain the link status information sent by the acceleration node;

a processing module, configured to generate a source routing table according to the link state information acquired by the transceiver module, where the source routing table includes a path from a source acceleration node to a destination acceleration node;

The transceiver module is used to obtain the first IP address of the destination end for which the user applies for the network acceleration service;

The processing module is further configured to generate a location routing table, where the location routing table includes the correspondence between the first IP address and the second acceleration node;

The transceiver module is further configured to send the location routing table and the source routing table corresponding to the first acceleration node to the first acceleration node, where the location routing table is used to guide the first acceleration node according to the first acceleration node. An IP address determines the second acceleration node, the first IP address is the destination address of the data request from the first terminal, and the source routing table is used for the first acceleration node to obtain a target path, and the target path is used to direct the data request to be forwarded to the second acceleration node, and the second acceleration node is used to forward the data to the destination end.
The central controller according to claim 22, wherein the processing module is further configured to: determine the second acceleration node according to the first IP address of the destination; establish the first IP address and the The corresponding relationship of the second acceleration node is generated; the location routing table is generated according to the corresponding relationship.
The central controller according to claim 23, wherein when the destination is a cloud area,

The processing module is further configured to determine the second acceleration node deployed in the cloud area according to the first IP address.
The central controller according to claim 23, wherein when the destination terminal is a terminal or a server, the processing module is further configured to query an IP address database according to the first IP address to determine the destination physical location of the terminal; determine the second acceleration node closest to the physical location.
The central controller according to any one of claims 22-25, characterized in that,

The first terminal is configured with an SDK plug-in, and the SDK plug-in is preset with the address information of the central controller;

The transceiver module is further configured to receive a request sent by the first terminal; and feed back the IP address of the first acceleration node to the first terminal according to the request, where the IP address of the first acceleration node is used for The first terminal sends the data request to the first acceleration node by using the SDK tunnel.
The central controller according to any one of claims 22-25, characterized in that,

The transceiver module is further configured to send traffic diversion information to the network management system, where the traffic diversion information includes IP information of the destination terminal, and the traffic diversion information is used to trigger the network management system to send an ACL policy instruction to the first network device, The second acceleration node is an acceleration node deployed in the first network device, and the ACL policy instruction is used to trigger the first network device to direct the data request from the first terminal to the first network device Speed up nodes.
The central controller of claim 22, wherein:

The transceiver module is further configured to acquire a mode parameter, where the mode parameter includes a first mode and a second mode, wherein the first mode is used to indicate that the destination of the network acceleration service is a cloud area, and the second mode The destination used to indicate the network acceleration service is the second terminal or the server.
A communication system, characterized by comprising a plurality of acceleration nodes as claimed in any one of claims 16 to 21 and a central controller as claimed in any one of claims 22 to 28, wherein the center The deployment environment of the controller belongs to the first cloud service provider, and the deployment environment of the multiple acceleration nodes belongs to the second cloud service provider, an application service provider or a telecommunication operator.
A central controller is characterized in that, comprises a processor, and the processor is coupled with at least one memory, and the processor is used to read the computer program stored in the at least one memory, so that the central controller executes the following steps: The method of any one of claims 9 to 15.
A computer program product, comprising computer program code, characterized in that, when the computer program code is executed by a computer, the computer is made to implement the method according to any one of claims 1 to 8 ; or, causing a computer to implement the above method as claimed in any one of claims 9 to 15.
A computer-readable storage medium, characterized in that it is used for storing computer programs or instructions, which, when executed, cause a computer to execute the method according to any one of claims 1 to 8; A computer implements the method of any one of claims 9 to 15 above.