WO2017161938A1 - Packet transmission method and device - Google Patents

Abstract

Disclosed are a packet transmission method and device. The method comprises: an aggregation agent receives a first packet of each of at least two virtual machines (VMs) from a virtualized network function (VNF) pool; the aggregation agent obtains a second packet according to the first packet, the second packet comprising a source Internet Protocol (IP) address and a destination IP address, the source IP address of the second packet being the IP address of the aggregation agent, the destination IP address of the second packet being the IP address of a server; the aggregation agent sends the second packet by means of aggregation links established between the aggregation agent and the server, the number of the aggregation links being less than that of sublinks. Therefore, the number of communication links established between a VNF pool and the server can be reduced, thereby saving the memory resource of the server.

Description

Message transmission method and device

The present application claims the priority of the Chinese Patent Application, filed on March 22, 2016, the application Serial No. in.

Technical field

The present invention relates to the field of communications technologies, and in particular, to a message transmission method and apparatus.

Background technique

NFV (Network Function Virtualization) technology replaces dedicated network element devices in communication networks with industry-standard x86 servers, storage devices, and switching devices, saving operators investment costs and helping operators With more and more flexible network capabilities, NFV technology is becoming more widely used.

The VNF (Virtual Network Function) runs on the NFV system in the form of software. The VNF is provided with a virtual machine (VM). Each VM can be thought of as a separate network functional entity that can be restarted independently and has its own root access, users, Internet Protocol (IP) addresses, memory, procedures, files, applications, system libraries, and Configuration files, etc.

In order to improve network performance, multiple VMs are deployed in the VNF. In the network function virtualization, multiple VMs deployed in the VNF form a VNF pool to implement load sharing and improve network reliability.

In the VNF Pool, if multiple VMs are started, each of the multiple VMs establishes a communication link with the server, such as a Transmission Control Protocol (TCP) connection, and performs packet transmission. If the number of VMs is large, excessive communication links are created between the VNF Pool and the server, which consumes more memory resources of the server.

Summary of the invention

The embodiment of the invention provides a message transmission method and device to reduce VNF Pool and server The number of communication links created between them saves the memory resources of the server.

In a first aspect, a packet transmission method is provided, in which a convergence agent receives a first packet from each of at least two VMs in a VNF Pool, and converts the first packet to obtain a first packet The second packet, the source IP address of the second packet is an IP address of the aggregation proxy, the destination IP address of the second packet is an IP address of the server, and then the convergence proxy passes the second packet The aggregation link between the aggregation agent and the server is sent to the server, so that the VNF pool externally presents the aggregation link. The server does not perceive the VNF pool as a whole and does not have multiple VMs under the VNF Pool. The number of the aggregation links is less than the number of the sub-links, which saves the number of communication links under the server, so as to save server memory resources.

In the embodiment of the present invention, the aggregation proxy may be load balancing (LB), and the LB may be built-in or externally placed in the VNF Pool. Specifically, the LB built in the VNF Pool or the LB externally placed in the VNF Pool may be determined according to the VNF Pool of the VM that establishes the sub-link with the aggregation proxy, if the VM that establishes the sub-link with the aggregation proxy belongs to a different one. The VNF Pool can be aggregated by using an external LB. If the VM that establishes a sub-link with the aggregation proxy belongs to a VNF Pool, the built-in LB can be used for aggregation.

The number of the aggregation links in the embodiment of the present invention may be established according to the number of the sub-links. If the number of sub-links is small, the transmission can be completed through one aggregation link, and only one aggregation link can be established. If the number of sub-links is too large and the load of the aggregation link is too heavy through an aggregation link, at least two aggregation links can be created.

The sub-link in the embodiment of the present invention refers to a link established between each VM in the VNF Pool and the aggregation proxy, and the at least two VMs are connected to the aggregation proxy through at least two sub-links, and The at least two VMs are in one-to-one correspondence with the at least two sub-links, and the sub-links are used to transmit the first packet. The source IP address of the first packet is the IP address of the VM, and the source port of the first packet is the port of the VM. The destination IP address of the first packet is the IP address of the aggregation proxy, and the destination port of the first packet. The IP address of the aggregation agent.

In the embodiment of the present invention, the aggregation link refers to a link between the aggregation proxy and the server, and the aggregation proxy and the server are connected by an aggregation link, and the aggregation link is used to transmit the second report. For example, the source IP address of the first packet is replaced by the IP address of the aggregation proxy, and the source IP address of the second packet is obtained, and the IP address of the server is used to replace the first packet. The destination IP address of the packet, and the destination IP address of the second packet.

In the embodiment of the present invention, the second packet includes a source IP address and a destination IP address, and the second packet may further include a source port and a destination port, where the source port of the first packet may be replaced by the port of the aggregation proxy. The source port of the second packet is obtained, that is, the source port of the second packet is the port of the aggregation proxy. The destination port of the second packet is replaced by the port of the server, and the destination port of the second packet is the port of the server.

The content included in the second packet in the embodiment of the present invention may be specifically determined according to the tuple information used in the second packet, where the tuple information refers to, for example, a five-tuple or a seven-tuple group. The information of the IP quintuple is preferably included in the second packet of the present invention, including the source port, the source IP address, the destination port, and the destination IP address. And the adopted agreement.

In the embodiment of the present invention, the second packet may include different content according to different transmission protocols.

In a possible design, the sub-link is a sub-TCP connection, and the aggregation link is a converged TCP connection. The second packet further includes a sequence number, where the sequence number of the second packet is a sequence number allocated by the aggregation proxy according to the aggregate TCP connection. In the embodiment of the present invention, the sequence number assigned by the aggregation link can be replaced by the sequence number assigned by the sub-link, and the sequence number of the second packet is obtained.

In the embodiment of the present invention, the aggregation proxy may send the second report to the server through the convergence TCP connection according to the source IP address, the source port, the destination IP address, the destination port, and the sequence number of the second packet included in the second packet. For example, the packets in the sub-TCP connection are aggregated and sent to the server in the aggregate TCP connection.

In the embodiment of the present invention, after the second aggregation packet is sent to the server by the convergence TCP connection, the convergence proxy may also receive the second response packet of the second packet by the server through the convergence TCP connection. The second response message includes the confirmation serial number of the second message, and the confirmation sequence of the second message The number is the sum of the sequence number of the second packet and the data length of the second packet; the aggregation agent searches for the pre-save according to the confirmation serial number of the second packet included in the second response packet. a relationship list, where the sequence number of the second packet and the data length of the second packet are determined, where the pre-saved relationship list includes the sequence number of the second packet and the data of the second packet. a length and a correspondence between the sub-TCP connections for transmitting the first packet; the aggregation proxy according to the determined sequence number of the second packet and the data length of the second packet, Determining a sub-TCP connection for transmitting the first packet in the relationship list; the aggregation proxy obtaining a first response packet of the first packet according to the second response packet, the first response packet The source IP address is the IP address of the aggregation proxy, and the destination IP address of the first response packet is the IP address of the VM to which the determined sub-TCP connection is connected, and the first response packet includes The confirmation serial number of the first message, the confirmation sequence of the first message The column number is the sum of the sequence number of the first packet and the data length of the first packet; the aggregation proxy is connected to the determined sub-TCP connection by using the determined sub-TCP connection. The VM forwards the first response packet to implement receiving and forwarding the response packet.

Optionally, the convergence TCP connection is a newly established TCP connection when the aggregation agent receives the first packet transmitted by the first sub-TCP connection in the VNF Pool, and receives the first transmission of the subsequent sub-TCP connection. When a packet is received, the establishment of a converged TCP connection is not triggered.

Through the above-mentioned possible implementation manners, the embodiment of the present invention implements the aggregation of the packets in the sub-TCP connection of each VM to a converged TCP connection by the aggregation proxy to transmit and receive the packets, thereby saving the TCP link on the server side. The number to save server memory resources.

In another possible design, the sub-link is a remote user dial-up authentication system RADIUS sub-session, the aggregation link is a RADIUS convergence session, and the second packet further includes a matching identifier, where the The matching identifier of the second packet is a matching identifier allocated by the aggregation proxy according to the RADIUS aggregation session. In the embodiment of the present invention, the matching identifier of the RADIUS sub-session is replaced by the matching identifier re-assigned by the RADIUS aggregation session, and the matching identifier of the second packet is obtained, that is, the matching identifier of the second packet is based on the RADIUS aggregation session. The allocation of information such as the IP address and port can avoid duplication of matching identifiers assigned with the RADIUS sub-session.

In the embodiment of the present invention, the aggregation proxy may send the second to the server through the RADIUS aggregation session according to the source IP address, the source port, the destination IP address, the destination port, and the matching identifier of the second packet included in the second packet. The packet is sent to the server in the RADIUS sub-session.

In the embodiment of the present invention, after the second aggregation packet is sent to the server by the RADIUS aggregation session, the convergence proxy may also receive the second response packet of the second packet by the server through the RADIUS convergence session. The second response packet includes a matching identifier of the second packet. The aggregation proxy searches for the pre-created proxy table according to the matching identifier of the second packet included in the second response packet, and determines the RADIUS sub-session. a matching identifier of the first packet and a RADIUS sub-session for transmitting the first packet, where the proxy table stores a matching identifier of the second packet, and a matching of the first packet An identifier and a correspondence between the RADIUS sub-sessions that transmit the first packet; the aggregation proxy obtains the first response packet of the first packet according to the second response packet, where the first The source IP address of the response packet is the IP address of the aggregation proxy, and the destination IP address of the first response packet is the IP address of the VM to which the determined RADIUS subsession is connected; Determining the number The matching identifier of a packet and the RADIUS sub-session transmitting the first packet forward the first response packet to the VM connected to the determined RADIUS sub-session.

Optionally, the RADIUS aggregation session is an established RADIUS session between the aggregation proxy and the server, and the RADIUS convergence session can be obtained by directly modifying the established RADIUS session.

Through the above-mentioned possible implementation manners, the embodiment of the present invention implements the aggregation of the packets in the RADIUS sub-sessions of different VMs into the RADIUS aggregation session through the aggregation proxy to send and receive the packets, which can save the number of RADIUS sessions on the server side. In order to save server memory resources.

In a further possible design, the aggregation agent releases the aggregation link in case the acknowledgment that the sub-link is released, to further save resources.

In a second aspect, a message transmission apparatus is provided, and the apparatus for implementing message transmission has a corresponding function for implementing the message transmission method according to the first aspect. The functions may be implemented by hardware or by corresponding software implemented by hardware. The hardware or software includes one or more modules corresponding to the functions described above. For example, the message transmission device includes a receiving unit and a processing unit, and the receiving unit is configured to receive a first message from each of at least two VMs in the VNF Pool, where the at least two VMs pass The at least two sub-links are connected to the packet transmission device, and the at least two VMs are in one-to-one correspondence with the at least two sub-links, and the sub-links are used to transmit the first packet. The processing unit is configured to obtain a second packet according to the first packet, where the second packet includes a source IP address and a destination IP address, and a source IP address of the second packet is the packet The IP address of the transmission device, and the destination IP address of the second packet is the IP address of the server. The sending unit is configured to send the second packet obtained by the processing unit by using an aggregation link established between the packet transmission device and the server, where the number of the aggregation links is smaller than The number of sub-links is reduced to save the number of communication links under the server, thereby saving server memory resources.

In a third aspect, a convergence agent is provided, the aggregation agent comprising a processor and a memory, wherein the memory stores a computer readable program, and the processor implements the first aspect by running a program in the memory Message transmission method.

In a fourth aspect, a computer storage medium is provided for storing the above-mentioned computer software instructions for implementing a message transmission device or a convergence agent, comprising a program for performing the message transmission method according to the above first aspect.

DRAWINGS

FIG. 1 is a network architecture diagram of a packet transmission method according to an embodiment of the present invention;

2 is a flowchart of implementing a message transmission method according to an embodiment of the present invention;

FIG. 3 is a schematic diagram of a process of message transmission according to an embodiment of the present disclosure;

4 is a schematic diagram of an implementation process of packet transmission performed by an LB built in a VNF Pool according to an embodiment of the present invention;

FIG. 5 is a schematic diagram showing an implementation process of message transmission by an LB placed in a VNF Pool according to an embodiment of the present invention; intention;

6 is a schematic diagram of a process of transmitting a packet by using a TCP protocol according to an embodiment of the present invention;

FIG. 7 is a schematic diagram of a process for transmitting a packet by using a RADIUS aggregation session according to an embodiment of the present invention;

FIG. 8 is a schematic structural diagram of a message transmission apparatus according to an embodiment of the present disclosure;

FIG. 9 is a schematic structural diagram of a convergence agent according to an embodiment of the present invention.

detailed description

The technical solutions in the embodiments of the present invention are described in detail below with reference to the accompanying drawings in the embodiments of the present invention.

The message transmission method provided by the embodiment of the present invention can be applied to the network architecture shown in FIG. 1. In Figure 1, each of the multiple VMs deployed in the VNF Pool establishes a communication link with the server and performs information interaction through the established communication link. For example, a communication link 1 is established between the VM1 and the server, and information is exchanged through the communication link 1. The communication link 2 is established between the VM2 and the server, and information is exchanged through the communication link 2, and the VM3 and the server are established. Communication link 3 and information exchange via communication link 3. In network function virtualization, multiple VMs (such as VM1, VM2, and VM3) deployed in the VNF Pool may start and interact with the server at the same time. In this case, the server needs to connect multiple communication links (for example) Communication link 1, communication link 2 and communication link 3) will consume more server memory resources.

It should be noted that, in the embodiment of the present invention, the network system architecture shown in FIG. 1 further includes load balancing (LB), and the LB mainly performs load sharing between VMs according to the load sharing policy, and VM1, VM2, and The packets exchanged between the VM3 and the server are transparently transmitted. In the network architecture shown in FIG. 1 , the LB is built in the VNF Pool as an example, but is not limited thereto. The LB may be built in or external to the VNF Pool.

It should be further noted that the structure of each VM involved in FIG. 1 is only for illustrative purposes, and is not limited thereto. For example, the VM may include a VM control unit (sControl) having a control function, and may also include VM forwarding unit (sForward) with forwarding function, the combination of each VM control unit and forwarding unit can be regarded as an independent network function entity, Figure 1 Not shown.

In the embodiment of the present invention, in order to avoid excessive communication links between the VM and the server, the memory resources of the server are consumed, and packets from multiple VMs in the VNF Pool can be aggregated and interacted with the server to save the server. Memory resources.

In the embodiment of the present invention, a packet of a plurality of VMs in the VNF Pool is aggregated and interacted with the server, and the aggregation proxy is implemented, and a communication link is established between the VM and the aggregation proxy, and the VM and the sub-link are connected. A correspondence. The VM sends the message sent to the server to the aggregation agent first through the communication link established between it and the aggregation agent. A communication link between the aggregation agent and the server is smaller than a communication link between the VM and the aggregation agent. The aggregation agent aggregates the received message and sends it to the server through the communication link established between the server and the server. . The VNF pool is configured to externally present the aggregation link. The server does not know that there are multiple VMs in the VNF pool. The virtual VMs in a VNF pool are actually virtualized into one pool, and the number of the aggregation links is smaller than the number of the aggregation links. The number of the sub-links saves the number of communication links under the server, so as to save server memory resources.

In the embodiment of the present invention, the link established between the VM and the aggregation proxy is referred to as a sub-link, and the packet transmitted by the sub-link is referred to as a first packet. The link established between the aggregation proxy and the server is called an aggregation link, and the packet transmitted by the aggregation link is called a second packet.

FIG. 2 is a flowchart of an implementation of a packet transmission method according to an embodiment of the present invention, and FIG. 3 is a schematic diagram of a packet transmission process according to an embodiment of the present invention.

S101: Each VM in the VNF Pool establishes a sub-link with the aggregation agent. As shown in FIG. 3, the VNF Pool includes VM1, VM2, and VM3, and VM1, VM2, and VM3 respectively establish sub-links with the aggregation agent, and are established. The sub-links are sub-link 1, sub-link 2, and sub-link 3, respectively.

S102: Create a new aggregation link between the aggregation proxy and the server, as shown in Figure 3.

In the embodiment of the present invention, the aggregation proxy can establish an aggregation link when receiving the packet transmitted through the first sub-link, and does not trigger the establishment of the aggregation link when receiving the packet transmitted through the subsequent sub-link. For example, in the embodiment of the present invention, when the VM1 requests to establish the sub-link 1 with the aggregation proxy, the trigger is triggered. Establish an aggregation link.

The number of the aggregation links established between the aggregation proxy and the server in the embodiment of the present invention may be determined according to the number of sub-links established between the VM and the aggregation proxy. If the number of the sub-links is small, the aggregation link may be adopted. Once the transfer is complete, only one aggregation link can be established. If the number of sub-links is too large and the load of the aggregation link is too heavy through an aggregation link, at least two aggregation links can be created. In the following embodiments of the present invention, the aggregation link is taken as an example, but is not limited thereto. FIG. 3 illustrates an example of including an aggregation link.

S103: Each VM of the multiple VMs in the VNF Pool respectively transmits a first packet by using a sub-link, and the aggregation proxy receives the first packet from each VM of the multiple VMs in the VNF Pool. For example, in FIG. 3, the first message from VM1 is transmitted through sub-link 1, the first message from VM2 is transmitted through sub-link 2, and the first message from VM3 is transmitted through sub-link 3.

S104: The convergence agent obtains the second packet according to the first packet.

In the embodiment of the present invention, the second packet is obtained according to the first packet, and is usually obtained by modifying the header of the first packet, for example, the following manner may be adopted: Obtaining, by the proxy, the packet header information of the first packet, where the packet header information includes the tuple information of the first packet in the sub-link, where the tuple information may be an N-tuple Information, where N is greater than or equal to 3, for example, more specific N-tuple information is triplet information or quintuple information or quaternary information. The information of the IP quintuple is included in the first packet, that is, the source port, the source IP address, the destination port, and the destination. The IP and the adopted protocol are described as examples, but are not limited. The aggregation agent replaces the tuple information of the first packet according to the tuple information of the second packet, obtains the packet header information of the second packet, and then obtains the second report by combining the payload of the first packet. Text.

The tuple information of the second packet in the embodiment of the present invention includes a source IP address and a destination IP address, and may further include a source port and a destination port, where the source port of the first packet may be replaced by the port of the aggregation proxy. The source port of the second packet is obtained, that is, the source port of the second packet is the port of the aggregation proxy. Replace the destination port of the first packet with the port of the server to obtain the destination of the second packet. The port, that is, the destination port of the second packet is a port of the server.

The specific content of the second packet in the embodiment of the present invention may be determined according to the tuple information used by the second packet, where the tuple information may be N-tuple information, where N is greater than or equal to 3, for example, more specific. The N-tuple information is triad information or quintuple information or quaternary group information. The information of the IP quintuple is preferably included in the second packet of the present invention, including the source port, the source IP address, the destination port, and the destination IP address. And the adopted agreement. The following describes the tuple information of the second packet, including the IP quintuple, as an example, but is not limited thereto.

S105: The convergence agent sends the second packet by using an aggregation link established between the proxy and the server.

In the embodiment of the present invention, the aggregation agent may add the first packet transmitted through the sub-link 1, the sub-link 2, and the sub-link 3 to the aggregation link according to the receiving sequence of the first packet, and implement the external server and the VNF. A small number of communication links are maintained between the pools, and fewer ports (sockets) are used for message transmission. For example, in the implementation process shown in FIG. 3, VM1, VM2, and VM3 respectively send the first packet to the aggregation proxy by using socket1, socket2, and socket3, and the aggregation proxy receives the first packet and obtains the second packet, and sends the second packet to the server through socket4. Second message.

Embodiments of the Present Invention The message transmission method according to the above embodiment will be described below in conjunction with practical applications.

In the embodiment of the present invention, the aggregation proxy may be a component that is deployed in the network architecture shown in FIG. 1 or may be a component that is integrated in the LB, which is not limited in the embodiment of the present invention. In the embodiment of the present invention, if the aggregation proxy is a component integrated in the LB, it can be understood as a function of adding an aggregation proxy to an existing LB with a load sharing function. .

The embodiment of the present invention is described below by taking the aggregation agent as a component integrated in the LB, that is, the aggregation proxy is an LB that adds a new aggregation proxy function.

The LB that adds the aggregation proxy function in the embodiment of the present invention may be built in or external to the VNF Pool. Specifically, the LB built in the VNF Pool or the LB externally placed in the VNF Pool can be used. It is determined according to the VNF Pool to which the VM that establishes the sub-link with the aggregation agent belongs. In a possible implementation, if the VMs that establish the sub-links with the aggregation proxy belong to different VNF Pools, the external LBs may be used for aggregation. In another possible implementation manner, if the VM that establishes the sub-link with the aggregation agent belongs to a VNF Pool, the built-in LB can be used for aggregation. It can be understood that, in the case that the VMs that establish the sub-links of the aggregation agent belong to the same VNF pool, the LBs may be used for aggregation, which is not limited in the embodiment of the present invention.

Figure 4 is a schematic diagram of an implementation process of an LB that is built in a VNF Pool to send packets through an aggregation link. In Figure 4, VM1, VM2, and VM3 form a VNF Pool, a sub-link 1 is established between VM1 and LB, a sub-link 2 is established between VM2 and LB, and a sub-link 3 is established between VM3 and LB. An aggregation link is established between the LB and the server. The number of aggregation links is shown in Figure 4, but it is not limited. The VM1 sends the first packet to the LB through the sub-link 1, the VM2 sends the first packet to the LB through the sub-link 2, and the VM3 sends the first packet to the LB through the sub-link 3, where VM1, VM2 and VM3 send The source IP address of the first packet is different, and the destination IP address is the same as the IP address of the LB. After receiving the first packet sent by VM1, VM2, and VM3, the LB replaces the source IP address of the first packet with the IP address of the LB, and replaces the destination IP address of the first packet with the IP address of the server, thereby obtaining the second packet. The packet, wherein the source IP address of the second packet is an IP address of the aggregation proxy, and the destination IP address of the second packet is an IP address of the server. The LB sends the second packet through an aggregation link established between the LB and the server.

Figure 5 is a schematic diagram of an implementation process of sending an LB through a convergence link to an LB that is externally placed in the VNF Pool. In Figure 5, VM1 and VM2 form a VNF Pool, and VM3 and VM4 form a VNF Pool. In the embodiment of the present invention, the form of the VM in each VNF Pool is not limited. In the embodiment of the present invention, VM1 and VM2 are mutually standby, and VM3 and VM4 are mutually active as an example. One of VM1 and VM2 establishes sub-link 1 with LB at the same time. One of VM3 and VM4 establishes sub-link 2 with LB at the same time. In Figure 5, VM1 and LB establish sub-link 1, VM3 and LB. The sublink 2 is established as an example for description. An aggregation link is established between the LB that is externally placed in the VNF Pool and the server. In Figure 5, the first packet is sent through the sub-link, and the LB processes the first packet to obtain the second packet, and passes through the convergence chain. The process of sending the second packet is different from the implementation process in FIG. 4, except that the VM belongs to a different VNF Pool, and the LB with the aggregation function is externally placed in the VNF Pool, and other similarities are not described herein.

The schematic diagram of the packet sending process shown in FIG. 4 and FIG. 5 shows that, in the embodiment of the present invention, multiple VMs that send the first packet may belong to the same VNF Pool, or may belong to different VNF Pools.

In the embodiment of the present invention, when the packet is transmitted, the content of the second packet is different according to the adopted transmission protocol. For example, if the Transmission Control Protocol (TCP) is used, the second packet is used. The serial number can also be included. If the User Datagram Protocol (UDP) is used, the second packet may further include a matching identifier. The following embodiments of the present invention are respectively exemplified for the foregoing two protocols.

First, the TCP protocol is taken as an example. When the TCP protocol is used for message transmission, the sub-link established between the VM and the LB can be called a sub-TCP connection. The aggregation link established between the LB and the server can be called Converged TCP connections.

FIG. 6 is a schematic diagram of a process for transmitting a message using the TCP protocol according to an embodiment of the present invention. In Figure 6, VM1, VM2, and VM3 in the VNF Pool establish a sub-TCP connection with the LB, respectively. The sub-TCP connection established by VM1 and LB is TCP1, and TCP1 is used to transmit the first message from VM1. The sub-TCP connection established by VM2 and LB is TCP2, and TCP2 is used to transmit the first message from VM2. The sub-TCP connection established by VM3 and LB is TCP3, and TCP3 is used to transmit the first packet from VM3. When the TCP protocol is used for message transmission, in order to ensure the sequence of transmitting messages and to ensure that all transmitted messages can be reassembled in the normal order, each sub-TCP connection must assign a sequence number to each transmitted message, in other words. The first packet further includes a sequence number, and the sequence number is in one-to-one correspondence with the sub-TCP connection, for example, the sequence number of the first packet transmitted by the TCP1 is seq1n, and the first packet transmitted by the TCP2 The sequence number is seq2n, and the sequence number of the first message transmitted through TCP3 is seq3n.

The first packet sent by VM1, VM2, and VM3 in the embodiment of the present invention includes an IP quintuple (source IP, source port, destination IP, destination port, and transport protocol), and includes a sequence number as an example. The correspondence between the IP quintuple, the serial number, and the sub-TCP connection is as shown in Table 1.

Table 1

In the embodiment of the present invention, the LB receives the first packet sent by VM1, VM2, and VM3, and obtains the second packet according to the first packet. The content included in the second packet in the embodiment of the present invention may be obtained according to the content included in the first packet. For example, the first packet includes an IP quintuple and a sequence number, and the second packet also includes an IP quintuple and a sequence number, except that the IP packet included in the second packet is five yuan. The group and serial number are different from the IP quintuple and serial number included in the first message. The IP quintuple of the second packet is obtained by replacing the IP quintuple of the first packet according to the convergence TCP connection, for example, replacing the source IP address of the first packet transmitted through TCP1, TCP2, and TCP3 by using the IP address of the LB. The address is the source IP address of the second packet, that is, the source IP address of the second packet is the IP address of the LB. The destination IP address of the second packet is obtained by using the IP address of the server to replace the destination IP address of the first packet transmitted by TCP1, TCP2, and TCP3. The destination IP address of the second packet is the IP address of the server. The source port of the first packet is replaced by the port of the LB, and the source port of the second packet is the port of the LB. The destination port of the second packet is obtained by using the port of the server to replace the destination port of the first packet transmitted by TCP1, TCP2, and TCP3, that is, the destination port of the second packet is the port of the server. TCP4 uses the same transmission protocol as TCP1, TCP2, and TCP3, that is, the transmission protocol of the second packet is TCP.

The sequence number of the second packet is obtained by replacing the sequence number of the first packet with the sequence number assigned by the convergence TCP connection. The second packet transmitted by the TCP4 in the embodiment of the present invention can be understood as a packet obtained by replacing the IP quintuple included in the packet header information of the first packet transmitted by the TCP1, the TCP2, and the TCP3. Therefore, the sequence numbers assigned by the TCP4 for the first message transmitted by TCP1, TCP2, and TCP3 when the sequence number is assigned are seq4n, seq4n+length1, and seq4n+length1+length2, respectively. among them, Length1 is the length of the first packet transmitted by TCP1, and length2 is the length of the first packet transmitted by TCP2. After the sequence number of the first packet is replaced by the sequence number assigned by the TCP4, the sequence number of the second packet is obtained, that is, the sequence number of the second packet obtained according to the first packet transmitted by the TCP1 is seq4n, and is transmitted according to the TCP2. The sequence number of the second packet obtained by the first packet is seq4n+length1, and the sequence number of the second packet obtained according to the first packet transmitted by TCP3 is seq4n+length1+length2.

The correspondence between the IP quintuple of the second packet and the serial number is shown in Table 2.

Table 2

In the embodiment of the present invention, after receiving the first packet sent by VM1, VM2, and VM3, the LB can process the first packet according to the foregoing conversion manner to obtain the second packet, and then obtain the second TCP connection. TCP4) sends the second message to the server.

In the embodiment of the present invention, after the LB sends the second packet to the server through the convergence TCP connection, the LB may also receive the response packet of the second packet, and obtain the first response packet according to the second packet. The response packet of the packet forwards the response packet of the first packet to VM1, VM2, and VM3. The following is a description of the present invention. The response packet of the second packet is referred to as a second response packet, and the response packet of the first packet is referred to as a first response packet.

Specifically, when the TCP transmission protocol is used for message transmission, the message sent by the sender can use the serial number to confirm whether the peer receives the message. When the sender sends the message, the sequence number of the message is placed in the message to be acknowledged. In the queue, the retransmission timer is started at the same time. If the response packet of the packet is received, the sequence number is confirmed, and the confirmation sequence number is used to determine that the packet has been received. If it is determined that the packet has been received, the packet is received from the queue. When the queue to be confirmed is deleted, when the TCP transmission protocol is used for message transmission, the serial number can also be used. Ensure that all transmitted messages can be reassembled in the normal order.

In the embodiment of the present invention, the LB aggregates the first packet transmitted by the TCP1, the TCP2, and the TCP3 to the TCP4, and sends the sequence number of the second packet allocated by the TCP4 to the to-be-confirmed queue of the TCP4. The serial number of the second message, the data length of the second message, and the correspondence between the sub-TCP connections (TCP1, TCP2, and TCP3) for transmitting the first message. The second response packet of the second packet includes an acknowledgement sequence number of the second packet, where the acknowledgement sequence number of the second packet is usually the sum of the sequence number of the second packet and the data length of the second packet. Therefore, the LB can determine the sequence number of the second packet and the data length of the second packet by using the confirmation sequence number of the second packet, and further determine the sub-TCP connection for transmitting the first packet, and determine the The sub-TCP connection of a packet forwards the response packet of the first packet to the corresponding VM.

In the embodiment of the present invention, the LB may pre-store the sequence number of the second packet, the data length of the second packet, and the correspondence between the sub-TCP connections (TCP1, TCP2, and TCP3) for transmitting the first packet, for example, Table 3 Shown.

table 3

Serial number of the second message	Data length of the second message	Sub-TCP connection
Seq4n	Length1	TCP1
Seq4n+length1	Length2	TCP2
Seq4n+length1+length2	Length3	TCP3

After receiving the second response packet, the LB searches for the pre-stored relationship list according to the confirmation sequence number of the second packet included in the second response packet, and determines the sequence number of the second packet and the second packet. a sequence number of the second packet equal to the acknowledgement sequence number of the second packet and a data length of the second packet, and then according to the sequence number of the second packet and the second report The data length of the text, and the sequence number of the second packet, the data length of the second packet, and the correspondence between the sub-TCP connections transmitting the first packet determine the sub-TCP connection for transmitting the first packet.

After determining the sub-TCP connection of the first packet to be transmitted in the embodiment of the present invention, the aggregation proxy obtains the first response packet of the first packet, where the source IP address of the first response packet is the convergence. Agent An IP address, the destination IP address of the first response packet is an IP address of the VM to which the determined sub-TCP connection is connected, and the first response packet includes an acknowledgement sequence number of the first packet, where the The confirmation sequence number of a message is usually the sum of the sequence number of the first message and the data length of the first message. The aggregation proxy forwards the first response packet to the VM connected to the determined sub-TCP connection by using the determined sub-TCP connection. After receiving the first response packet, the VM may determine, according to the acknowledgement sequence number of the first packet included in the first response packet, the first packet that is sent by the first response packet.

In the embodiment of the present invention, the aggregation proxy can replace the IP quintuple of the second response packet and the second quintuple of the second response packet by determining the IP quintuple of the first packet transmitted by the sub-TCP connection and the acknowledgment sequence number of the first packet. The serial number of the packet, and the IP quintuple of the first response packet of the first packet and the sequence number.

The IP quintuple of the second response message and the sequence number of the second message in the embodiment of the present invention may be as shown in Table 4 below.

Table 4

The IP quintuple of the first response message and the sequence number of the first packet in the embodiment of the present invention may be as shown in Table 5 below.

table 5

It should be noted that length1 in Table 5 of the embodiment of the present invention is the data length of the first packet sent by VM1, length2 is the data length of the first packet sent by VM2, and length3 is the data length of the first packet sent by VM3. .

The embodiment of the present invention implements the LB to aggregate the packets transmitted through the sub-TCP connection to a converged TCP connection for sending and receiving packets, thereby saving the number of TCP connections on the server side, thereby saving the server. The purpose of memory resources.

Optionally, in the embodiment of the present invention, if all the sub-TCP connections in the VNF Pool are released, to further save resources, the newly created converged TCP connection may also be released.

Optionally, in the embodiment of the present invention, the LB transmits the packet to a converged TCP connection, so the VNF Pool presents a converged connection, so if the number of VMs in the VNF Pool increases or decreases, the LB will not be affected. The number of converged TCP connections with the server, that is, the server does not perceive the number of VMs in the VNF Pool as a whole, and realizes that the VMs in the VNF Pool are truly virtualized into one pool, forming a clouded resource pool. And can make the VM free from geographical restrictions.

The Remote Authentication Dial In User Service (RADIUS) is a widely used method for transmitting data by UDP. Therefore, the RADIUS session is used as an example for the process of transmitting packets by using the UDP protocol in the embodiment of the present invention. In the embodiment of the present invention, the sub-link established between the VM and the LB is a RADIUS sub-session, and the aggregation link established between the LB and the server is a RADIUS aggregation session.

The implementation process of packet transmission using RADIUS session and reporting by TCP transmission protocol The implementation process of the text transmission is similar, and the similarities are not described here. The following only explains the differences.

When a packet is transmitted through a RADIUS session, the first packet transmitted by the RADIUS sub-session can be converted into a second packet, and then transmitted through a RADIUS aggregation session. For example, the RADIUS1, RADIUS2, and RADIUS3 can be transmitted in Figure 7. After a message is converted into a second packet, it is transmitted through a RADIUS4. Different from the TCP transmission protocol, when a RADIUS session is used for packet transmission, the LB does not need to add an aggregation link. You can modify the packet in the established RADIUS session.

In the message in the RDIUS session, the matching identifier (Identifier) is used to match the request message and the response message, and the matching identifier is generally a sequentially increasing number, so in order to avoid the matching identifier and the VM in the aggregation link. The matching identifiers in the sub-links are repeated. In the embodiment of the present invention, each of the VM sub-links may be assigned a new matching identifier in the aggregation link.

FIG. 7 is a schematic diagram of a process of transmitting a message through a RADIUS session according to an embodiment of the present invention. In Figure 7, VM1, VM2, and VM3 in the VNF Pool establish a RADIUS subsession with the LB, respectively. The RADIUS sub-session established by VM1 and LB is RADIUS1, and RADIUS1 is used to transmit the first packet from VM1. The RADIUS sub-session established by VM2 and LB is RADIUS2, and RADIUS2 is used to transmit the first packet from VM2. The RADIUS sub-session established by VM3 and LB is RADIUS3, and RADIUS3 is used to transmit the first packet from VM3. RADIUS1, RADIUS2, and RADIUS3 assign the matching identifier of the first packet to the first packet they transmit.

In the embodiment of the present invention, the first packet sent by VM1, VM2, and VM3 includes an IP quintuple as an example, and the IP quintuple, the matching identifier, and the RADIUS subsession included in the first packet are used as an example. The correspondence between them is shown in Table 6.

Table 6

The LB replaces RADIUS1 with the IP quintuple of the second packet transmitted by RADIUS4.

The IP quintuple of the first packet transmitted by RADIUS2 and RADIUS3 uses RADIUS4 to replace the matching identifier of the first packet with the matching identifier of the second packet. The second packet is obtained. The correspondence between IP quintuple, matching identifier, and RADIUS subsession is shown in Table 7:

Table 7

The LB sends the second packet through a RADIUS aggregation session established between the LB and the server according to the IP quintuple of the second packet and the matching identifier.

In the embodiment of the present invention, after the LB sends the second packet to the server through the RADIUS aggregation session, the LB may also receive the response packet of the second packet, and obtain the first response packet according to the second packet. The response packet of the packet forwards the response packet of the first packet to VM1, VM2, and VM3.

In the embodiment of the present invention, the LB may pre-create a proxy table, where the proxy table stores a matching identifier of the second packet, a matching identifier of the first packet, and a RADIUS for transmitting the first packet. The correspondence between sub-sessions. After the LB receives the second response packet of the second packet through the RADIUS convergence session, the LB may be based on the second response because the second response packet includes a matching identifier of the second packet. The matching identifier of the second packet included in the packet searches for the pre-created proxy table, and obtains a matching identifier of the first packet allocated by the RADIUS sub-session and a RADIUS sub-session for transmitting the first packet. The aggregation proxy obtains the first response packet of the first packet, where the source IP address of the first response packet is an IP address of the aggregation proxy, and the first response packet is The destination IP address is the determined IP address of the VM to which the RADIUS sub-session is connected; the aggregation proxy determines the matching identifier according to the obtained first packet and the RADIUS sub-session for transmitting the first packet. The VM connected to the RADIUS sub-session forwards the first response packet.

In the embodiment of the present invention, the first response packet may be replaced by the IP quintuple of the first packet and the matching identifier of the first packet transmitted by the RADIUS sub-session, and the IP quintuple of the second response packet and the matching identifier are replaced. get.

The correspondence between the IP quintuple and the matching identifier in the second response packet in the embodiment of the present invention is as follows:

Table 8

The IP quintuple and the matching identifier of the first response message in the embodiment of the present invention may be as shown in Table 9 below:

Table 9

It can be understood that the number of matching identifiers in the RADIUS is less than 256. Therefore, in a specific implementation, the LB can process the RADIUS sub-sessions of multiple VMs through a queue or a caching mechanism, if the VM If the number of RADIUS sub-sessions exceeds 256, it can be used by releasing the completed RADIUS sub-session, or by waiting for an extra matching identifier.

The foregoing embodiment of the present invention implements the LB to aggregate the packets transmitted by the RADIUS sub-sessions of multiple VMs in the VNF Pool to the RADIUS aggregation session for sending and receiving, which can save the number of RADIUS sessions on the server side. In order to save server memory resources.

Based on the message transmission method provided by the foregoing embodiment, the embodiment of the present invention further provides a message transmission apparatus 100. FIG. 8 is a schematic structural diagram of a packet transmission apparatus 100 according to an embodiment of the present invention. As shown in FIG. 8, a packet transmission apparatus 100 according to an embodiment of the present invention includes a receiving unit 101, a processing unit 102, and a sending unit 103, where

The receiving unit 101 is configured to receive a first packet from each of at least two VMs in the VNF Pool, where the at least two VMs are connected to the packet transmission device by using at least two sub-links, and The at least two VMs are in one-to-one correspondence with the at least two sub-links, and the sub-links are used to transmit the first packet.

The processing unit 102 is configured to obtain a second packet according to the first packet, where the second packet includes a source Internet Protocol IP address and a destination IP address, and a source IP address of the second packet is The IP address of the message transmission device 100, and the destination IP address of the second message is the IP address of the server.

The sending unit 103 is configured to send the second packet obtained by the processing unit 102 by using an aggregation link established between the packet transmission device 100 and the server, where the convergence link is The number is less than the number of the sub-links.

Optionally, the second packet further includes a source port and a destination port, where the source port of the second packet is a port of the packet transmission device 100, and a destination port of the second packet The port for the server.

In a possible implementation, the sub-link is a sub-transmission control protocol TCP connection, the aggregation link is a converged TCP connection, and the second packet further includes a sequence number; wherein the second report The serial number of the text is the serial number assigned by the processing unit 102 according to the aggregate TCP connection.

Optionally, the receiving unit 101 is further configured to: send, by the sending unit 103, the first obtained by the processing unit 102 by using a convergence TCP connection established between the packet transmission device 100 and the server. Receiving, by the convergence TCP connection, the second response message that is sent by the server to the second packet by using the convergence TCP connection, where the second response packet includes a confirmation sequence number, where the confirmation sequence number is The sum of the sequence number of the second message and the data length of the second message.

The processing unit 102 is further configured to: search for a pre-saved relationship list according to the confirmation sequence number included in the second response message, and determine a sequence number of the second packet and data of the second packet. a length, according to the determined sequence number of the second packet and the data length of the second packet, determining, in the relationship list, a sub-TCP connection for transmitting the first packet, where the pre-save is performed The relationship list includes a sequence number of the second packet, a data length of the second packet, and a correspondence between sub-TCP connections for transmitting the first packet. Obtaining a first response packet of the first packet according to the second response packet, where a source IP address of the first response packet is an IP address of the packet transmission apparatus 100, and the first response The destination IP address of the packet is the IP address of the VM connected to the determined sub-TCP connection, and the first response packet includes the confirmation sequence number of the first packet, and the confirmation sequence number of the first packet is The sum of the sequence number of the first message and the data length of the first message.

The sending unit 103 is further configured to: forward, by using the determined sub-TCP connection by the processing unit 102, the first response packet to the VM connected to the determined sub-TCP connection.

Optionally, the aggregation link is a newly established TCP link when the processing unit 102 receives the packet of the first VM sub-link in the VNF Pool.

In another possible implementation manner, the sub-link is a remote user dial-up authentication system RADIUS sub-session, the aggregation link is a RADIUS convergence session, and the second packet further includes a matching identifier; The matching identifier of the second packet is a matching identifier allocated by the processing unit 102 according to the RADIUS convergence session.

In another possible implementation manner, after the sending unit 103 sends the second packet obtained by the processing unit 102 by using a RADIUS convergence session established between the packet transmission device 100 and the server, Receiving, by the RADIUS aggregation session, the server response to the first And a second response message of the second message, where the second response message includes a matching identifier of the second message.

The processing unit 102 is further configured to: search a pre-created proxy table according to a matching identifier of the second packet included in the second response packet, and determine a matching identifier of the first packet allocated by the RADIUS sub-session. And a RADIUS sub-session for transmitting the first packet, where the proxy table stores a matching identifier of the second packet, a matching identifier of the first packet, and a RADIUS that transmits the first packet. The correspondence between sub-sessions. Obtaining a first response packet of the first packet according to the second response packet, where a source IP address of the first response packet is an IP address of the packet transmission device 100, and the first response packet The destination IP address is the IP address of the VM to which the determined RADIUS subsession is connected;

The sending unit 103 is further configured to: connect to the determined RADIUS sub-session according to the determined matching identifier of the first packet and a RADIUS sub-session that transmits the first packet The VM forwards the first response message.

Optionally, the RADIUS convergence session is an established RADIUS session between the packet transmission device 100 and the server.

Optionally, the number of the aggregation links is established by the processing unit 102 according to the number of the sub-links.

Optionally, the message transmission device 100 is a load sharing LB built in or external to the VNF Pool.

Optionally, the processing unit 102 is further configured to: when the acknowledgment that the VM sub-links are released, release the aggregation link.

In the embodiment of the present invention, the message transmission device 100 aggregates the received packets from the VNF Pool that are transmitted through the multiple sub-links to the aggregation link, so that the VNF Pool presents the aggregation link and the number of the aggregation links. The number of sub-links is less than the number of sub-links. The server treats the entire VNF pool as a whole. The server does not know that there are multiple VMs in the VNF Pool. The virtual VMs in a VNF Pool can be virtualized into one pool, saving the number of communication links under the server. To achieve the purpose of saving resources.

In a specific implementation, the foregoing message transmission apparatus 100 of the embodiment of the present invention may be a convergence agent. 9 is a schematic structural diagram of a convergence proxy 200 according to an embodiment of the present invention. As shown in FIG. 9, the convergence proxy 200 adopts a general computer system structure, including a bus, a processor 201, a memory 202, and a communication interface 203, to implement the solution of the present invention. The program code is stored in memory 202 and is controlled by processor 201 for execution.

The bus can include a path to transfer information between various components of the computer.

The processor 201 can be a general purpose central processing unit (CPU), a microprocessor, an application-specific integrated circuit (ASIC), or one or more integrated circuits for controlling the execution of the program of the present invention. One or more memories included in the computer system, which may be read-only memory (ROM) or other types of static storage devices that can store static information and instructions, random access memory (RAM) Or other types of dynamic storage devices that can store information and instructions, or disk storage. These memories are connected to the processor via a bus.

The communication interface 203 can use devices such as any transceiver to communicate with other devices or communication networks, such as Ethernet, Radio Access Network (RAN), Wireless Local Area Network (WLAN), and the like.

A memory 202, such as a RAM, holds an operating system and a program for executing the inventive arrangements. The operating system is a program that controls the running of other programs and manages system resources.

The program stored in the memory 202 is used by the instruction processor 201 to perform the message transmission method according to the foregoing embodiment of the present invention, comprising: receiving a first message from each of at least two VMs in the VNF Pool, the at least Two VMs are connected to the aggregation proxy through at least two sub-links, and the at least two VMs are in one-to-one correspondence with the at least two sub-links, and the sub-links are used to transmit the first packet; Obtaining a second packet according to the first packet, where the second packet includes a source Internet protocol IP address and a destination IP address, and a source IP address of the second packet is an IP address of the aggregation proxy. The destination IP address of the second packet is the IP address of the server; the second packet is sent by the aggregation link established between the server and the server, where the number of the aggregation links is smaller than the sub-chain The number of roads achieves the goal of saving server memory resources.

It can be understood that the convergence agent 200 of this embodiment can be used to implement the foregoing method embodiments. For the specific implementation process, reference may be made to the related description of the foregoing method embodiments, and details are not described herein again.

The embodiment of the present invention further provides a computer storage medium for storing the computer software instructions used in the message transmission apparatus described in FIG. 8 or FIG. 9, which includes a program for executing the foregoing method embodiments. By executing the stored procedure, the packets in the VM sub-links in the received VNF Pool are aggregated into one aggregation link for transmission.

It should be noted that, in the embodiment of the present invention, “a plurality” means two or more. "and/or", describing the association relationship of the associated objects, indicating that there may be three relationships, for example, A and/or B, which may indicate that there are three cases where A exists separately, A and B exist at the same time, and B exists separately. The character "/" generally indicates that the contextual object is an "or" relationship.

It is apparent that those skilled in the art can make various modifications and variations to the invention without departing from the spirit and scope of the invention. Thus, it is intended that the present invention cover the modifications and modifications of the invention

Claims (22)

1. A message transmission method, comprising:

   The aggregation agent receives a first message from each of the at least two virtual machine VMs in the virtual network function pool VNF Pool, the at least two VMs being connected to the aggregation agent through at least two sub-links, and the At least two VMs are in one-to-one correspondence with the at least two sub-links, and the sub-links are used to transmit the first packet;

   The aggregation agent obtains a second packet according to the first packet, where the second packet includes a source Internet protocol IP address and a destination IP address, and a source IP address of the second packet is the aggregation proxy. An IP address, where the destination IP address of the second packet is an IP address of the server;

   The aggregation agent sends the second packet by using an aggregation link established between the aggregation agent and the server, where the number of the aggregation links is smaller than the number of the sub-links.
2. The method according to claim 1, wherein the second packet further includes a source port and a destination port;

   The source port of the second packet is a port of the aggregation proxy, and the destination port of the second packet is a port of the server.
3. The method according to claim 1 or 2, wherein the sub-link is a sub-transmission control protocol TCP connection, and the aggregation link is a converged TCP connection;

   The second message further includes a serial number;

   The sequence number of the second packet is a sequence number allocated by the aggregation proxy according to the convergence TCP connection.
4. The method of claim 3, wherein the method further comprises: after the second agent sends the second message through the aggregated TCP connection established between the server and the server, the method further comprising:

   The aggregation proxy receives the second response packet of the second packet in response to the second TCP packet, and the second response packet includes a confirmation sequence number of the second packet. The confirmation sequence number of the second message is the sum of the sequence number of the second message and the data length of the second message;

   The aggregation agent searches the pre-stored relationship list according to the confirmation sequence number of the second packet included in the second response packet, and determines the sequence number of the second packet and the data length of the second packet. The pre-stored relationship list includes a sequence number of the second packet, a data length of the second packet, and a correspondence between sub-TCP connections for transmitting the first packet.

   The aggregation agent determines, according to the determined sequence number of the second packet and the data length of the second packet, a sub-TCP connection for transmitting the first packet in the relationship list;

   The aggregation proxy obtains the first response packet of the first packet according to the second response packet, where the source IP address of the first response packet is an IP address of the aggregation proxy, and the first The destination IP address of the response packet is the IP address of the VM connected to the determined sub-TCP connection, and the first response packet includes an acknowledgement sequence number of the first packet, and the acknowledgement of the first packet The serial number is a sum of a sequence number of the first packet and a data length of the first packet;

   The aggregation proxy forwards the first response packet to the VM connected to the determined sub-TCP connection by using the determined sub-TCP connection.
5. The method according to claim 3 or 4, wherein the converged TCP connection is a newly established TCP connection when the aggregation agent receives the first packet transmitted by the first sub-TCP connection in the VNF Pool.
6. The method according to claim 1 or 2, wherein the sub-link is a remote user dial-up authentication system RADIUS sub-session, and the aggregation link is a RADIUS convergence session;

   The second message further includes a matching identifier;

   The matching identifier of the second packet is a matching identifier allocated by the aggregation proxy according to the RADIUS convergence session.
7. The method of claim 6, wherein the method further comprises: after the convening agent sends the second packet through a RADIUS aggregation session established between the server and the server, the method further comprising:

   The aggregation proxy receives, by the RADIUS aggregation session, a second response packet that is sent by the server to the second packet, where the second response packet includes a matching identifier of the second packet.

   The aggregation agent checks the matching identifier of the second packet included in the second response packet. Finding a pre-created proxy table, determining a matching identifier of the first packet allocated by the RADIUS sub-session, and a RADIUS sub-session transmitting the first packet, where the proxy table saves the matching of the second packet An identifier, a matching identifier of the first packet, and a correspondence between RADIUS sub-sessions transmitting the first packet;

   The aggregation proxy obtains the first response packet of the first packet according to the second response packet, where the source IP address of the first response packet is an IP address of the aggregation proxy, and the first The destination IP address of the response packet is the IP address of the VM to which the determined RADIUS subsession is connected;

   The aggregation agent forwards the first response to the VM connected to the determined RADIUS sub-session according to the determined matching identifier of the first packet and the RADIUS sub-session transmitting the first packet. Message.
8. The method according to claim 6 or 7, wherein the RADIUS aggregation session is an established RADIUS session between the aggregation proxy and the server.
9. The method according to any one of claims 1 to 8, wherein the number of the aggregation links is established according to the number of the sub-links.
10. The method according to any one of claims 1 to 9, wherein the aggregation agent is a load sharing LB built in or external to the VNF Pool.
11. The method of any of claims 1 to 10, further comprising:

    The aggregation agent releases the aggregation link in case the acknowledgment that the sub-links are released.
12. A message transmission device, comprising:

    a receiving unit, configured to receive a first packet from each of at least two virtual machine VMs in a virtual network function pool VNF Pool, the at least two VMs passing through at least two sub-links and the message transmission device Connecting, and the at least two VMs are in one-to-one correspondence with the at least two sub-links, where the sub-link is used to transmit the first packet;

    a processing unit, configured to obtain a second packet according to the first packet, where the second packet includes a source Internet Protocol IP address and a destination IP address, and a source IP address of the second packet is the packet The IP address of the transmission device, and the destination IP address of the second packet is an IP address of the server;

    a sending unit, configured to send the second packet obtained by the processing unit by using an aggregation link established between the packet transmission device and the server, where the number of the aggregation links is smaller than the sub- The number of links.
13. The device according to claim 12, wherein the second packet further includes a source port and a destination port;

    The source port of the second packet is a port of the packet transmission device, and the destination port of the second packet is a port of the server.
14. The device according to claim 12 or 13, wherein the sub-link is a sub-transmission control protocol TCP connection, and the aggregation link is a converged TCP connection;

    The second message further includes a serial number;

    The sequence number of the second packet is a sequence number allocated by the processing unit according to the convergence TCP connection.
15. The device according to claim 14, wherein the receiving unit is further configured to:

    After the sending unit sends the second packet obtained by the processing unit by using the convergence TCP connection established between the packet transmission device and the server, receiving the server response by using the convergence TCP connection. a second response packet of the second packet, where the second response packet includes an acknowledgement sequence number of the second packet, and the acknowledgement sequence number of the second packet is a sequence number of the second packet And a sum of data lengths of the second message;

    The processing unit is further configured to:

    Determining, according to the confirmation sequence number of the second packet included in the second response packet, a pre-stored relationship list, determining a sequence number of the second packet and a data length of the second packet, according to the Determining a sequence number of the second packet and a data length of the second packet, determining, in the relationship list, a sub-TCP connection for transmitting the first packet, the pre-saved relationship list And including a sequence number of the second packet, a data length of the second packet, and a correspondence between sub-TCP connections for transmitting the first packet;

    Obtaining a first response packet of the first packet according to the second response packet, where the first The source IP address of the packet is the IP address of the packet transmission device, and the destination IP address of the first response packet is the IP address of the VM to which the determined sub-TCP connection is connected, the first The response packet includes an acknowledgement sequence number of the first packet, where the acknowledgement sequence number of the first packet is a sum of a sequence number of the first packet and a data length of the first packet;

    The sending unit is further configured to:

    And the first response packet is forwarded to the VM connected to the determined sub-TCP connection by using the sub-TCP connection determined by the processing unit.
16. The device according to claim 14 or 15, wherein the aggregation link is a newly created TCP when the processing unit receives the message of the first VM sub-link in the VNF Pool by the receiving unit. link.
17. The device according to claim 12 or 13, wherein the sub-link is a remote user dial-up authentication system RADIUS sub-session, and the aggregation link is a RADIUS convergence session;

    The second message further includes a matching identifier;

    The matching identifier of the second packet is a matching identifier allocated by the processing unit according to the RADIUS aggregation session.
18. The device according to claim 17, wherein the receiving unit is further configured to:

    After the sending unit sends the second packet obtained by the processing unit by using a RADIUS aggregation session established between the packet transmission device and the server, receiving the server response through the RADIUS convergence session. a second response packet of the second packet, where the second response packet includes a matching identifier of the second packet;

    The processing unit is further configured to:

    Searching a pre-created proxy table according to the matching identifier of the second packet included in the second response packet, determining a matching identifier of the first packet allocated by the RADIUS sub-session, and transmitting the first packet a RADIUS sub-session, the proxy table storing a matching identifier of the second packet, a matching identifier of the first packet, and a correspondence between RADIUS sub-sessions transmitting the first packet;

    And obtaining, by using the second response packet, the first response packet of the first packet, where a source IP address of the first response packet is an IP address of the packet transmission device, and the first response packet The destination IP address of the text is the IP address of the VM to which the determined RADIUS subsession is connected;

    The sending unit is further configured to:

    Determining, according to the determined matching identifier of the first packet and the RADIUS sub-session transmitting the first packet, the first response packet to the VM connected to the determined RADIUS sub-session.
19. The device according to claim 17 or 18, wherein the RADIUS aggregation session is an established RADIUS session between the message transmission device and the server.
20. The apparatus according to any one of claims 12 to 19, wherein the number of the aggregation links is established by the processing unit according to the number of the sub-links.
21. The device according to any one of claims 12 to 20, wherein the message transmission device is a load sharing LB built in or external to the VNF Pool.
22. The device according to any one of claims 12 to 21, wherein the processing unit is further configured to:

    The aggregation link is released in the case of confirming that each VM sub-link is released.

Images