CN103095579B - TRILL network interconnected method, Apparatus and system - Google Patents

Abstract

The invention discloses a TRILL network interconnection method, device, and system, which relate to the field of communication technology and can solve the problem that each edge RB needs to allocate a large amount of hardware resources to support packet encapsulation and decapsulation of packets when data center DCs are interconnected. The problem of low document forwarding efficiency. The method of the present invention includes: the first edge routing bridge RB of the first data center DC receives the RB identification information sent by the second edge RB of the second DC, and the RB identification information carries the root RB in the second DC The combination of the RB identity ID and the second DC identity ID; the first edge RB establishes a distribution tree forwarding entry according to the RB identity information, so as to send a message according to the distribution tree forwarding entry. The present invention is mainly used in the process of interconnection between DCs.

Description

TRILL network interconnection method, device and system

technical field

The present invention relates to the field of communication technology, in particular to a TRILL network interconnection method, device and system.

Background technique

With the development of network technologies, the scale and number of data centers (Data Center, DC for short) are increasing rapidly. Generally, a DC is interpreted as a "multifunctional building that can accommodate multiple servers and communication equipment. These equipment are placed together because they have the same environmental requirements and physical security needs, and they are placed so that they are easy to maintain. .” So, simply put, a DC consists of multiple servers, switches, and routers. A Bridge running the Transparent Interconnection of Lots of Links (TRILL) protocol is called a Route-Bridge (RB), which is a bridge device with routing and forwarding characteristics. The network built by RB is called It is TRILL network (TRILL Campus).

In the prior art, the interconnection between DCs based on TRILL is implemented in the following two ways:

1. Use multiple DCs as a TRILLCampus. The egress RB of each DC manages the routing table for message forwarding between DCs.

2. Each DC is interconnected as an independent TRILLCampus. If the first host of the first DC needs to broadcast in a certain distribution tree of the second DC, the first RB connected to the first host in the first DC adds the first TRILL header to the data frame, and the first The RB in the DC forwards the data frame to the RB serving as a router in the first DC through the distribution tree in the first DC according to the information of the first TRILL message header, and the RB removes the first TRILL message header and sends The message is routed to the RB serving as a router of the second DC, and the RB adds a second TRILL message header to the message for forwarding to the root node of the distribution tree through the distribution tree in the second DC.

In the process of realizing the above-mentioned network interconnection, the inventor found that at least the following problems existed in the prior art:

In prior art 1, due to the limited processing capability and storage space of the router, the number of entries in the routing table managed by the router is limited, so the routing range is also limited, which cannot satisfy large-scale inter-DC interconnection. Because the router manages the routing paths between multiple DCs, the routing table is huge, so its convergence time will increase compared with when the router manages the routing table in one DC. Therefore, the existing technology is only suitable for small-scale inter-DC interconnection, and cannot support large-scale inter-DC interconnection.

In prior art 2, since each DC uses its own distribution tree to forward the message, when the message is forwarded between different DCs, the egress RB of the DC that sends the message needs to decapsulate the message, and the DC that receives the message The router needs to re-encapsulate the packet. The router needs to allocate a large amount of hardware resources to support packet encapsulation and decapsulation, and the efficiency of packet forwarding is low.

Contents of the invention

Embodiments of the present invention provide a TRILL network interconnection method, device and system, which can solve the problems of low message forwarding efficiency and inability to support large-scale inter-DC interconnection caused by encapsulating and decapsulating messages.

In a first aspect, the present invention provides a multi-link translucent interconnection TRILL network interconnection method, the method comprising:

The first edge RB of the first data center DC receives the RB identification information sent by the second edge RB of the second DC, and the RB identification information carries the RB identity ID of the root RB in the second DC and the A combination of the second DC identity ID;

The first edge RB establishes a distribution tree forwarding entry according to the RB identification information, so as to send packets according to the distribution tree forwarding entry.

In a first possible implementation manner of the first aspect, the first edge RB receives the RB identification information sent by the second edge RB through a Border Gateway Protocol (BGP).

In the first aspect or the first possible implementation of the first aspect, a second possible implementation of the first aspect is also provided, and in the second possible implementation of the first aspect In an implementation manner, the first edge RB establishes the distribution tree forwarding entry through the shortest path first algorithm (SPF) algorithm according to the RB identification information;

The first edge RB forwards the RB identification information to other RBs in the first DC, so that the other RBs establish their own distribution tree forwarding entries through the SPF algorithm according to the RB identification information.

In the first aspect or the first possible or second possible implementation of the first aspect, a third possible implementation of the first aspect is also provided, in the first aspect In a third possible implementation manner of , the first edge RB obtains the port number corresponding to the next-hop RB, and adds the port number corresponding to the next-hop RB to the distribution tree forwarding entry; the The other RBs obtain the port number corresponding to the next-hop RB, and add the port number corresponding to the next-hop RB to the distribution tree forwarding entry established by itself.

In the first aspect or the first possible, second possible or third possible implementation of the first aspect, a fourth possible implementation of the first aspect is also provided, in In the fourth possible implementation manner of the first aspect, the RB identification information sent by the first edge RB of the first data center DC to the second edge RB of the second DC further carries:

A virtual local area network VLAN ID, where the VLAN ID is used to identify the VLAN to which the root RB in the first DC and the root RB in the second DC belong; the first edge RB receives the message sent by the second edge RB. extended information, the extended information is used to describe the correspondence between the RB identification information of the root RB in the second DC and the VLAN ID.

In the first aspect or the first possibility, the second possibility, the third possibility or the fourth possible implementation of the first aspect, a fifth possibility of the first aspect is also provided In the fifth possible implementation of the first aspect, the first edge RB establishes the RB identification information of the root RB in the first DC and the tree root RB in the second DC according to the VLANID. The corresponding relationship of the RB identification information of the root RB, thereby converting the extended information into a local distribution tree forwarding entry.

In a second aspect, the present invention provides a multi-link translucent interconnection TRILL network interconnection method, the method comprising:

The second edge RB of the second data center DC sends RB identification information to the first edge RB of the first DC, and the RB identification information carries the RB identity ID of the root RB in the second DC and the first A combination of two DC IDs, so that the first edge RB establishes a distribution tree forwarding entry according to the RB identification information, and sends a message according to the distribution tree forwarding entry.

In a first possible implementation manner of the second aspect, the second edge RB sends the RB identification information to the first edge RB through a Border Gateway Protocol (BGP).

In the second aspect or the first possible implementation of the second aspect, a second possible implementation of the second aspect is also provided, and in the second possible implementation of the second aspect In an implementation manner, the RB identification information sent by the second edge RB to the first edge RB also carries:

A virtual local area network VLAN ID, the VLAN ID is used to identify the VLAN to which the root RB in the first DC and the root RB in the second DC belong; the second edge RB sends a message to the first edge RB Extended information, where the extended information is used to describe the correspondence between the RB identification information of the root RB in the second DC and the VLAN ID.

In a third aspect, the present invention provides a multi-link translucent interconnection TRILL network interconnection device, the device is a first edge RB in a first data center DC, and the device includes:

The receiving unit is configured to receive RB identification information sent by the second edge RB of the second DC, where the RB identification information carries the RB identity ID of the root RB in the second DC and the second DC identity A combination of IDs;

A processing unit, configured to establish a distribution tree forwarding entry according to the RB identification information received by the receiving unit.

In a first possible implementation manner of the third aspect, the receiving unit is further configured to receive the RB identification information sent by the second edge RB through a Border Gateway Protocol (BGP).

In the third aspect or the first possible implementation of the third aspect, a second possible implementation of the third aspect is also provided, and in the second possible implementation of the third aspect In an implementation manner, the processing unit specifically includes:

A calculation subunit, configured to establish the distribution tree forwarding entry through the shortest path SPF algorithm according to the RB identification information received by the receiving unit;

The sending subunit is configured to forward the RB identification information received by the receiving unit to other RBs in the first DC.

In the third aspect or the first possible or second possible implementation of the third aspect, a third possible implementation of the third aspect is also provided, in the third aspect In a third possible implementation manner of , the calculating subunit is further configured to obtain the port number corresponding to the next-hop RB, and add the port number corresponding to the next-hop RB to the distribution tree forwarding entry.

In the third aspect or the first possible, second possible or third possible implementation of the third aspect, a fourth possible implementation of the third aspect is also provided, in In a fourth possible implementation manner of the third aspect, the receiving unit is further configured to receive extended information sent by the second edge RB, where the extended information is used to describe the root RB in the second DC The corresponding relationship between the RB identification information and the VLAN ID.

In the third aspect or the first possibility, the second possibility, the third possibility or the fourth possible implementation of the third aspect, a fifth possibility of the third aspect is also provided In a fifth possible implementation of the third aspect, the processing unit is further configured to establish the RB identification information of the root RB in the first DC and the RB identification information of the root RB in the second DC according to the VLANID. Correspondence between RB identification information of the root RB.

In a fourth aspect, the present invention provides a multi-link translucent interconnection TRILL network interconnection device, the device is a second edge RB in a second data center DC, and the device includes:

A sending unit, configured to send RB identification information to the first edge RB of the first DC, where the RB identification information carries the RB identity ID of the root RB in the second DC and the second DC identity ID The combination.

In a first possible implementation manner of the fourth aspect, the sending unit is further configured to send the RB identification information to the first edge RB through a Border Gateway Protocol (BGP).

In the fourth aspect or the first possible implementation manner of the fourth aspect, a second possible implementation manner of the fourth aspect is also provided, and in the second possible implementation manner of the fourth aspect In an implementation manner of , the sending unit is further configured to send extended information to the first edge RB, where the extended information is used to describe the correspondence between the RB identification information of the root RB in the second DC and the VLANID .

In a fifth aspect, the present invention provides a multi-link semi-transparent interconnection TRILL network interconnection system, the system is composed of a first edge RB and a second edge RB.

In the TRILL network interconnection method, device and system provided by the present invention, since the RB identification information received by the first edge RB is a combination of the RB identity ID of the root RB in the second DC and the second DC identity ID, therefore When the first edge RB establishes a distribution tree forwarding entry according to the RB identification information, the distribution tree forwarding entry established by the first edge RB includes the distribution tree forwarding entry for each tree root RB inside the first DC In addition, there are distribution tree forwarding entries for each tree root RB in the second DC. Compared with the need to decapsulate and encapsulate the message in the first edge RB and the second edge RB in the prior art 2, the first edge RB and the second edge RB in the present invention do not need to decapsulate and encapsulate the message , which can reduce the allocation of hardware resources supporting packet encapsulation and decapsulation, and the packet forwarding efficiency is high. In addition, compared with prior art 1 where the first edge RB or the second edge RB manages all the routing tables in the first DC and the second DC, the first edge RB and the second edge RB manage the routing tables of their respective DCs respectively , so it can support large-scale interconnection between DCs.

Description of drawings

In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings that need to be used in the description of the embodiments or the prior art. Obviously, the accompanying drawings in the following description are only These are some embodiments of the present invention. Those skilled in the art can also obtain other drawings based on these drawings without creative work.

Fig. 1 is the method flowchart of a kind of TRILL network interconnection method in the embodiment of the present invention;

Fig. 2 is the method flowchart of another TRILL network interconnection method in the embodiment of the present invention;

Fig. 3 is a method flowchart of another TRILL network interconnection method in an embodiment of the present invention;

Fig. 4 is the method flowchart of another TRILL network interconnection method in the embodiment of the present invention;

FIG. 5 is a schematic structural diagram of a first first edge RB in an embodiment of the present invention;

FIG. 6 is a schematic structural diagram of a second first edge RB in an embodiment of the present invention;

FIG. 7 is a schematic structural diagram of a third first edge RB in an embodiment of the present invention;

FIG. 8 is a schematic structural diagram of the first second edge RB in an embodiment of the present invention;

9 is a schematic structural diagram of the first TRILL network interconnection system in an embodiment of the present invention;

FIG. 10 is a schematic structural diagram of a fourth first edge RB in an embodiment of the present invention;

FIG. 11 is a schematic structural diagram of a second second edge RB in an embodiment of the present invention;

FIG. 12 is a schematic structural diagram of a second TRILL network interconnection system in an embodiment of the present invention.

detailed description

The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only some of the embodiments of the present invention, not all of them. Based on the embodiments of the present invention, all other embodiments obtained by persons of ordinary skill in the art without creative efforts fall within the protection scope of the present invention.

In order to facilitate the understanding of the present invention, the structure of the TRILL message header is shown in Table 1.

Table 1

In Table 1: TRILLEthertype is the network type of the TRILL protocol; V is the TRILL version number, which is currently 0, and if it is not 0, the message will be discarded; R is a reserved field; M is the multicast identifier, 0 means unicast, 1 Indicates multicast; OpLng (0p-Length) is the length of the TRILL message header extension option; Hop is the hop number; EgressRBridgeNickname (egress RB name) is the RB identification information of the RB connected to the server if it is unicast, and if it is multicast It is the root RB identification information of the distribution tree; IngressRBridgeNickname (entry RB name) is the RB identification information of the RB connected to the server, and is used to identify the source RB that adds the TRILL message header to the message.

The naming of the tree root RB, other RBs, and next-hop RBs mentioned in the following embodiments is named according to the function of a certain RB in the process of sending a single TRILL message. The sending of a single TRILL message is based on the entry The RB name (the RB identification information of the first RB connected to the server) is the starting point, and the egress RB name is the ending point. Therefore, the same RB may serve as the tree root RB or other RBs or next-hop RBs in different message sending processes.

Embodiment one

The embodiment of the present invention provides a TRILL network interconnection method, as shown in Figure 1, the method includes:

Step 101, the first edge RB of the first data center DC receives RB identification information sent by the second edge RB of the second DC.

Wherein, the RB identity information carries a combination of the RB identity ID of the root RB in the second DC and the second DC identity ID.

The structure of the RB identification information is: DCID: RBID. By combining the DCID and the RBID, each RB in the first DC and the second DC has unique identification information. Optionally, the structure of the above RB identification information may also be: RBID: DCID.

Step 102, the first edge RB establishes a distribution tree forwarding entry according to the RB identification information.

The first edge RB establishes a distribution tree forwarding entry according to the RB identification information, so as to send packets according to the distribution tree forwarding entry. The first edge RB can send the received RB identification information to the RBs in the first DC, and the first edge RB and each RB in the first DC can establish the RB identification information for the first DC in addition to the RB identification information in the first DC. In addition to the distribution tree forwarding entries, distribution tree forwarding entries for each tree root RB in the second DC are also established according to the received RB identification information in the second DC.

Specifically, after receiving the RB identification information, the first edge RB generates a topology for the RB identification information in its own link state database (LinkStateDatabase, LSDB for short), including the following two methods:

1. Copy the topology of the first edge RB itself.

The first edge RB adds an entry in the LSDB, and the entry is a set of corresponding relationships after replacing the RB identifier of the first edge RB with the RB identifier information. For example: the RB identification information of the first edge RB is 01:01, and the entries starting with the first edge RB already exist in the LSDB as "01:01-01:02" and "01:01-01: 03". After the first edge RB receives the RB identification information sent by the second edge RB, if the RB identification information received at this time is 02:02, add entries: "02:02-01:02" and "02: 02-01:03".

2. Copy the topology of a root RB in the first DC.

The tree root can be selected randomly or according to preset rules. Select a root RB in the first DC, and then find the root RB as the starting entry in the LSDB, for example: select the root RB whose RB ID is 01:05, and start with 01:05 The entries are "01:05-01:03" and "01:05-01:06". After the first edge RB receives the RB identification information sent by the second edge RB, if the RB identification information received at this time is 02:02, add entries: "02:02-01:03" and "02: 02-01:06".

The preset rule may be but not limited to a hash algorithm. After the first edge RB adds an entry to the LSDB, the distribution tree forwarding entry can be formed by combining the entry in the LSDB with the SPF algorithm. When establishing a distribution tree forwarding entry, it is necessary to refer to the priority of each entry in the LSDB. The higher the priority of the starting RB identifier in the entry, the greater the probability of being determined as the tree root RB. Therefore, in order to ensure that the RB identification information of the root RB in the second DC is elected as the root in the first DC, the RB identification information of the root RB in the second DC and the RB identification information of the root RB in the first DC are assigned The same tree root priority (rootpriority).

After the first edge RB generates a topology for the RB identification information in its own LSDB, the first edge RB notifies the newly added entries in its own LSDB, that is, LSPs (topology), in the first DC.

When the second edge RB of the second data center DC receives the RB identification information sent by the first edge RB of the first DC, the second edge RB of the second DC can obtain the RB identification information of each tree root RB in the first DC , and then the interconnection between the first DC and the second DC can be realized.

If the first host of the first DC needs to broadcast in a certain distribution tree of the second DC, the first RB connected to the first host in the first DC adds a TRILL header to the data frame, and the TRILL The entry RB name in the message header is the name of the first RB connected to the first host in the first DC, and the exit RB name is the RB identification information of the RB that is the root of the distribution tree in the second DC. According to the information in the TRILL packet header, the RB in the first DC forwards the data frame to the first edge RB of the first DC through the RB in the first DC, and the first edge RB directly routes the packet to the second The second edge RB of the DC, after the second edge RB receives the message, each RB in the second DC checks the "M" field in the TRILL message header, and multicasts the message according to the distribution tree entry.

The above message sending process is only a one-way technical solution after the second edge RB of the second DC sends RB identification information to the first edge RB of the first DC, and the first edge RB of the first DC sends the second DC After the second edge RB of the first DC sends the RB identification information, and the first edge RB of the first DC sends the RB identification information to other edge RBs of other DCs, network interconnection between multiple DCs can be realized.

In the TRILL network interconnection method provided by the present invention, since the RB identification information received by the first edge RB is a combination of the RB ID of the root RB in the second DC and the ID of the second DC, when the first edge RB When an edge RB establishes a distribution tree forwarding entry according to the RB identification information, the distribution tree forwarding entry established by the first edge RB contains, in addition to the distribution tree forwarding entries for each tree root RB inside the first DC, The forwarding entry for each distribution tree root RB in the second DC. Compared with the need to decapsulate and encapsulate the message in the first edge RB and the second edge RB in the prior art 2, the first edge RB and the second edge RB in the present invention do not need to decapsulate and encapsulate the message , which can reduce the allocation of hardware resources supporting packet encapsulation and decapsulation, and the packet forwarding efficiency is high. In addition, compared with the first edge RB or the second edge RB in the prior art to manage all the routing tables in the first DC and the second DC, in the present invention, the first edge RB and the second edge RB respectively manage the respective DCs routing table, so it can support large-scale interconnection between DCs.

Embodiment two

The embodiment of the present invention provides a TRILL network interconnection method, as shown in Figure 2, the method includes:

Step 201, the second edge RB of the second data center DC sends RB identification information to the first edge RB of the first DC.

Wherein, the RB identity information carries a combination of the RB identity ID of the root RB in the second DC and the second DC identity ID, so that the first edge RB can establish The distribution tree forwarding entry is used to send the message according to the distribution tree forwarding entry.

In the TRILL network interconnection method provided by the embodiment of the present invention, the RB identification information sent from the second edge RB of the second DC to the first edge RB of the first DC can be used to distinguish RBs in the first DC from RBs in the second DC , so that each RB in the first DC and in the second DC has unique identification information. Compared with prior art 2 where the first edge RB and the second edge RB need to decapsulate and encapsulate the message, in the embodiment of the present invention, the first edge RB can receive the RB identification information in the second DC Establish a distribution tree forwarding table item, so that the first edge RB and the second edge RB do not need to decapsulate and encapsulate the message, reduce the allocation of hardware resources used by the edge RB to support message encapsulation and decapsulation, and improve the message forwarding efficiency high. In addition, compared with prior art 1 where the first edge RB or the second edge RB manages all the routing tables in the first DC and the second DC, in the embodiment of the present invention the first edge RB and the second edge RB manage their respective The routing table of the DC where it is located, so it can support large-scale interconnection between DCs.

Embodiment Three

As a detailed description and further expansion of Embodiment 1 and Embodiment 2, the embodiment of the present invention also provides a TRILL network interconnection method, as shown in FIG. 3 , the method includes:

Step 301, the second edge RB of the second data center DC sends RB identification information to the first edge RB of the first DC.

Wherein, the RB identity information carries a combination of the RB identity ID of the root RB in the second DC and the second DC identity ID, so that the first edge RB can establish The distribution tree forwarding entry is used to send the message according to the distribution tree forwarding entry.

The RB identification information sent by the second edge RB of the second DC to the first edge RB of the first DC can distinguish the RBs in the first DC and the RBs in the second DC, so that the RBs in the first DC and the RBs in the second DC have a unique identification information.

Specifically, the second edge RB sends the RB identification information to the first edge RB through a border gateway protocol BGP.

When the second edge RB sends the RB identification information to the first edge RB through BGP, in addition to sending the same information as the BGP content type in the prior art, it also sends the first additional information shown in Table 2, the first The additional information is used to indicate whether the RB corresponding to the RB identification information in the additional information is the root RB of the distribution tree. The first additional information also includes the RB identification information. Optionally, the VLAN ID corresponding to the RB identification information is also included in the first additional information, for matching the data format in the first DC.

Table 2 (first additional information)

Yes or No (whether it is the root RB) Root Nickname (RB identification information of the root RB)

After the second edge RB of the second data center DC sends the RB identification information to the first edge RB of the first DC, the first edge RB of the first data center DC receives the RB identification information sent by the second edge RB of the second DC. Wherein, the RB identity information carries a combination of the RB identity ID of the root RB in the second DC and the second DC identity ID. Through the first entry in the first additional information, it can be determined whether the RB identification information is the RB identification information of the root RB.

Specifically, the first edge RB receives the RB identification information sent by the second edge RB through Border Gateway Protocol BGP.

Step 302, the first edge RB establishes a distribution tree forwarding entry according to the RB identification information.

The first edge RB establishes a distribution tree forwarding entry according to the RB identification information, so as to send packets according to the distribution tree forwarding entry. The first edge RB can send the received RB identification information to other RBs in the first DC, and the first edge RB and other RBs in the first DC can establish the RB identification information for the first DC in addition to the RB identification information in the first DC. In addition to the distribution tree forwarding entries, distribution tree forwarding entries for each tree root RB in the second DC are also established according to the received RB identification information in the second DC.

Specifically, as shown in FIG. 4, the step 302 specifically includes:

Step 401, the first edge RB establishes the distribution tree forwarding entry through the shortest path first algorithm (SPF) algorithm according to the RB identification information.

After receiving the RB identification information, the first edge RB generates a topology for the RB identification information in its own LSDB, including the following two methods:

1. Copy the topology of the first edge RB itself.

The first edge RB adds an entry in the LSDB, and the entry is a set of corresponding relationships after replacing the RB identifier of the first edge RB with the RB identifier information. For example: the RB identification information of the first edge RB is 01:01, and the entries starting with the first edge RB already exist in the LSDB as "01:01-01:02" and "01:01-01: 03". After the first edge RB receives the RB identification information sent by the second edge RB, if the RB identification information received at this time is 02:02, add entries: "02:02-01:02" and "02: 02-01:03".

2. Copy the topology of a root RB in the first DC.

The tree root can be selected randomly or according to preset rules. Select a root RB in the first DC, and then find the root RB as the starting entry in the LSDB, for example: select the root RB whose RB ID is 01:05, and start with 01:05 The entries are "01:05-01:03" and "01:05-01:06". After the first edge RB receives the RB identification information sent by the second edge RB, if the RB identification information received at this time is 02:02, add entries: "02:02-01:03" and "02: 02-01:06".

The preset rule may be but not limited to a hash algorithm. After the first edge RB adds an entry to the LSDB, the distribution tree forwarding entry can be formed by combining the entry in the LSDB with the SPF algorithm. When establishing a distribution tree forwarding entry, it is necessary to refer to the priority of each entry in the LSDB. The higher the priority of the starting RB identifier in the entry, the greater the probability of being determined as the tree root RB. Therefore, in order to ensure that the RB identification information of the root RB in the second DC is elected as the root in the first DC, the RB identification information of the root RB in the second DC and the RB identification information of the root RB in the first DC are assigned The same tree root priority (rootpriority).

The SPF algorithm uses a certain RB as the root (ROOT) to calculate the distance to each destination RB. When calculating, each RB will calculate the topology structure diagram of the routing domain according to a unified database. The structure diagram is similar to a tree. In the SPF algorithm, it is called the shortest path tree. After the operation of the SPF algorithm, a path with the least number of hops between the root RB and the destination RB of the shortest path tree can be obtained, and the RB identification information of the RB closest to the root RB in the path is used as the distribution tree forwarding table item to save. In addition, the port list calculated by the SPF algorithm as the port number on the root RB can be saved as a distribution tree forwarding entry, that is, the first edge RB obtains the port number of the corresponding next-hop RB, and The port number corresponding to the next-hop RB is added to the distribution tree forwarding entry. Messages can be sent to multiple RBs at the same time through the port number.

Step 402, the first edge RB forwards the RB identification information to other RBs in the first DC.

After the first edge RB generates a topology for the received RB identification information in its own LSDB, the first edge RB notifies the newly added entries in its own LSDB, that is, LSPs (topology), in the first DC. After the announcement, each RB in the first DC can obtain the newly added entry corresponding to the RB identification information.

The first edge RB forwards the RB identification information to other RBs in the first DC, so that the other RBs establish their own distribution tree forwarding entries through the SPF algorithm according to the RB identification information. In order to enable all RBs in the first DC to send packets to RBs in the second DC, the first edge RB sends the received RB identification information to other RBs in the first DC. The other RBs can obtain the RB identification information of the next-hop RB for sending messages to the RB in the second DC through the SPF algorithm described in step 401, and save the RB identification information. In addition, the port list calculated by the SPF algorithm as the port number on the root RB is stored as a distribution tree forwarding entry, and the port number can be used to send messages to multiple destination RBs at the same time, that is, the other RBs obtain the corresponding The port number of the next-hop RB, adding the port number of the corresponding next-hop RB to the distribution tree forwarding entry established by itself.

If the first host of the first DC needs to broadcast in a certain distribution tree of the second DC, the first RB connected to the first host in the first DC adds a TRILL header to the data frame, and the TRILL The entry RB name in the message header is the name of the first RB connected to the first host in the first DC, and the exit RB name is the name of the RB serving as the root RB of the distribution tree in the second DC. The first RB connected to the first host compares the egress RB name in the TRILL header with each distribution tree forwarding entry saved by itself, and finds the RB identification information of the next-hop RB corresponding to the egress RB name or port list, and send the message to the next hop RB, each RB checks the "M" field in the TRILL message header, and multicasts the message according to the distribution tree entry;

If the egress RB name is the RB identification information of a root RB inside the first DC, the message is a multicast message inside the first DC.

Compared with prior art 2 where the first edge RB and the second edge RB need to decapsulate and encapsulate the message, the above message sending process passes through the second edge RB of the second DC to the first edge of the first DC The RB identification information sent by the RB can distinguish the RBs in the first DC and the RBs in the second DC, so that each RB in the first DC and in the second DC has unique identification information. The first edge RB can establish a distribution tree forwarding entry according to the received RB identification information in the second DC, and at the same time, the first edge RB sends the received RB identification information to other RBs in the first DC, so that the first Other RBs in the DC establish distribution tree forwarding entries according to the received RB identification information in the second DC. Thus, under the premise that the first edge RB and the second edge RB do not need to decapsulate and encapsulate the message, the message can be successfully sent to the RB of another DC, so that the RB can check the TRILL message header In the "M" field, the message is multicast according to the distribution tree entry, which reduces the allocation of hardware resources used by the edge RB to support message encapsulation and decapsulation, and the message forwarding efficiency is high. In addition, compared with prior art 1 where the first edge RB or the second edge RB manages all the routing tables in the first DC and the second DC, the first edge RB and the second edge RB respectively manage The routing table of their respective DCs, so it can support large-scale interconnection between DCs.

Further, the RB identification information sent by the second edge RB to the first edge RB also carries: a virtual local area network (VLAN) ID, and the VLAN ID is used to identify the root RB in the first DC A VLAN to which the root RB in the second DC belongs.

At this time, step 301 is further refined as, the second edge RB of the second data center DC sends RB identification information to the first edge RB of the first DC, specifically including: the second edge RB sends the first edge RB The edge RB sends extended information, where the extended information is used to describe the correspondence between the RB identification information of the root RB in the second DC and the VLAN ID.

When the second edge RB sends the RB identification information to the first edge RB through BGP, in addition to sending the same content type information as in the prior art, it also sends the second additional information shown in Table 3, the second The additional information is used to indicate whether it is the root RB, the RB identification information of the root RB, and the correspondence between the root RB and the VLAN. Wherein, the RB identification information of the root RB and the corresponding relationship between the root RB and the VLAN are extended information.

Table 3 (second additional information)

Yes or No (whether it is the root RB) Root Nickname (RB identification information of the root RB) Extended information: Ethernet Tag ID (VLAN ID)

At this time, the first edge RB of the first data center DC receives the RB identification information sent by the second edge RB of the second DC, specifically including: the first edge RB receives the extended information sent by the second edge RB , the extended information is used to describe the correspondence between the RB identification information of the root RB in the second DC and the VLAN ID.

The first edge RB of the first data center DC receives the RB identification information sent by the second edge RB of the second DC and also carries: a virtual local area network VLANID, and the VLANID is used to identify the RB and the RB in the first DC. The VLAN to which the root RB in the second DC belongs.

Step 302 is further refined as, the first edge RB establishes the corresponding relationship between the RB identification information of the RB in the first DC and the RB identification information of the root RB in the second DC according to the VLANID, thereby the Extended information is translated into local distribution tree forwarding table entries.

After the first edge RB of the first DC receives the second additional information sent by the second edge RB of the second DC, it searches for the RB identification information of the RB corresponding to the VLANID in the first DC according to the extended information therein, that is, the VLANID , and store the RB identification information, the RB identification information of the root RB and the VLAN ID in the second additional information as a distribution tree forwarding entry.

Each distribution tree forwarding table item constitutes a distribution tree forwarding table, as shown in Table 4:

Table 4 (distribution tree forwarding table)

Second RB identification information VLAN ID First RB identification information 02:01 VLAN1 01:01 02:03 VLAN2 01:02 ... ... ...

Wherein, the second RB identification information is the RB identification information of the root RB in the second DC, and the first RB identification information is the RB identification information corresponding to the RB in the same VLAN in the first DC and the second RB identification information. The edge RB of the first DC in Table 4 can use the RB identification information in the second DC in the TRILL header to find out the RB identification information corresponding to the RB identification information of the root RB in the first DC based on the VLAN.

For example, if the second host of the second DC needs to broadcast in a distribution tree of the first DC, the first RB connected to the second host in the second DC adds a TRILL header to the data frame, so The entry RB name in the TRILL message header is the name of the first RB connected to the second host in the second DC, and the exit RB name is the RB identification information of the root RB of the distribution tree in the first DC. If the name of the egress RB in the header of the TRILL message is 02:01 and the corresponding VLAN ID is LVAN1 in the RB identification information in the second DC, then the first edge RB of the first DC obtains the RB identification in the first DC through lookup table 4 Information 01:01, the first edge RB of the first DC decapsulates and encapsulates the TRILL packet header, and changes the name of the egress RB in the TRILL packet header to 01:01, so as to be forwarded inside the first DC. Since the forwarding of the message within the first DC is an existing technology, it can be successfully sent to the RB corresponding to the egress RB name, and each RB checks the "M" field in the TRILL message header, and forwards the message according to the distribution tree entry. multicast.

Compared with the decapsulation and encapsulation of the message by the first edge RB of the first DC and the second edge RB of the second DC in the prior art 2, the above process of sending the message saves the decapsulation of the second edge RB step, at the same time, other RBs except the edge RB do not need to calculate the next-hop distribution tree forwarding table items for RBs in other DCs, reducing the allocation of hardware resources used by edge RBs to support packet encapsulation and decapsulation, Packet forwarding efficiency is high, further reducing the workload of other RBs. In addition, compared with prior art 1 where the first edge RB or the second edge RB manages all the routing tables in the first DC and the second DC, the first edge RB and the second edge RB respectively Manage the routing table of their respective DCs, so it can support large-scale interconnection between DCs.

Further, step 302 may be further refined as, the first edge RB randomly establishes a distribution tree forwarding entry according to the received RB identification information. The random establishment is that the first edge RB takes the received RB identification information in the second DC and any tree root RB identification information in the first DC as a set of correspondences. Since any RB in the first DC can send a message to the VLAN corresponding to the VLANID through a certain port of the RB, the technical effect of interconnection between different DCs can also be achieved through the above method.

Further, in addition to sending messages between different DCs through each RB, the establishment of the above-mentioned distribution tree forwarding table items and the maintenance of the distribution tree forwarding table items can also be completed through a controller (Controller) configured separately for each DC . Each RB only needs to receive instruction information sent by the controller, such as sending a message through a certain interface. Since the functions of the controller and each edge RB and other RBs are clear, the working efficiency of each RB will be further improved.

The TRILL network interconnection method provided by the embodiment of the present invention is a one-way technical solution after the second edge RB of the second DC sends RB identification information to the first edge RB of the first DC. After the second edge RB of the second DC sends the RB identification information, and the first edge RB of the first DC sends the RB identification information to other edge RBs of other DCs, network interconnection among multiple DCs can be realized.

In an implementation of the TRILL network interconnection method provided by the embodiment of the present invention, the RB identification information sent by the second edge RB of the second DC to the first edge RB of the first DC can distinguish the RB in the first DC from the second edge RB. The RBs in the DC make each RB in the first DC and the RBs in the second DC have unique identification information. The first edge RB can establish a distribution tree forwarding entry according to the received RB identification information in the second DC, and at the same time, the first edge RB sends the received RB identification information to other RBs in the first DC, so that the first Other RBs in the DC establish distribution tree forwarding entries according to the received RB identification information in the second DC. Thus, under the premise that the first edge RB and the second edge RB do not need to decapsulate and encapsulate the message, the message can be successfully sent to the RB of another DC, so that the RB can check the TRILL message header In the "M" field, the message is multicast according to the distribution tree entry, which reduces the allocation of hardware resources used by the edge RB to support message encapsulation and decapsulation, and the message forwarding efficiency is high.

Another implementation of the TRILL network interconnection method provided by the embodiment of the present invention is compared with the decapsulation and encapsulation of the message by the first edge RB of the first DC and the second edge RB of the second DC in the second prior art , the sending process of the above message omits the decapsulation step of the second edge RB, and at the same time, other RBs other than the edge RBs of each DC do not need to calculate the next-hop distribution tree forwarding table items for the RBs in other DCs, reducing The small edge RB is used to support the allocation of hardware resources for packet encapsulation and decapsulation, and the packet forwarding efficiency is high, further reducing the workload of other RBs.

In addition, compared with prior art 1 where the first edge RB or the second edge RB manages all the routing tables in the first DC and the second DC, in the above two implementations the first edge RB and the second edge RB manage The routing table of their respective DCs, so it can support large-scale interconnection between DCs.

Embodiment four

An embodiment of the present invention provides a TRILL network interconnection device. The device is a first edge RB in a first data center DC. As shown in FIG. 5, the first edge routing bridge RB includes:

The receiving unit 51 is configured to receive RB identification information sent by the second edge RB of the second DC.

Wherein, the RB identity information carries a combination of the RB identity ID of the root RB in the second DC and the second DC identity ID, so that the first edge RB can establish The distribution tree forwarding entry is used to send the message according to the distribution tree forwarding entry.

Specifically, the receiving unit 51 is further configured to receive the RB identification information sent by the second edge RB through the Border Gateway Protocol BGP.

When receiving the RB identification information, the receiving unit 51 receives the first additional information in addition to the information of the same type as the BGP content in the prior art. The first additional information is used to indicate whether the RB corresponding to the RB identification information in the additional information is the root RB of the distribution tree, and the first additional information also includes the RB identification information. Optionally, the VLAN ID corresponding to the RB identification information is also included in the first additional information, for matching the data format in the first DC.

The processing unit 52 is configured to establish a distribution tree forwarding entry according to the RB identification information received by the receiving unit 51 .

In addition to establishing the distribution tree forwarding entry for the first DC according to the RB identification information in the first DC, the processing unit 52 also establishes a distribution tree forwarding entry for each tree in the second DC according to the RB identification information in the second DC received by the receiving unit 51. The distribution tree forwarding entry of the root RB.

The processing unit 52, as shown in FIG. 6 , specifically includes:

The calculation subunit 521 is configured to establish the distribution tree forwarding entry through the shortest path SPF algorithm according to the RB identification information received by the receiving unit 51 .

The SPF algorithm uses a certain RB as the root (ROOT) to calculate the distance to each destination RB. When calculating, each RB will calculate the topology structure diagram of the routing domain according to a unified database. The structure diagram is similar to a tree. In the SPF algorithm, it is called the shortest path tree. The calculation subunit 521 can obtain a path with the least number of hops between the root RB and the destination RB as the shortest path tree through the operation of the SPF algorithm, and use the RB identification information closest to the root RB in the path as the distribution tree The forwarding entry is saved to the storage unit 61 . In addition, the storage unit 61 can also store the port list composed of port numbers on the root RB calculated by the calculation subunit 521 through the SPF algorithm as a distribution tree forwarding entry, so that the sending subunit 522 can simultaneously send Multiple destination RBs send packets. That is, the first edge RB obtains the port number of the corresponding next-hop RB, and adds the port number of the corresponding next-hop RB to the distribution tree forwarding entry.

The sending subunit 522 is configured to forward the RB identification information received by the receiving unit 51 to other RBs in the first DC.

In order to enable each RB in the first DC to send a message to an RB in the second DC, the sending subunit 522 sends the RB identification information received by the receiving unit 51 to other RBs in the first DC. Other RBs can obtain the RB identification information of the next-hop RB for sending messages to the second intra-DC RB through the calculation subunit 521 therein, and save the RB identification information to the storage unit 61 .

The calculation subunit 521 is further configured to obtain the port number of the corresponding next-hop RB, and add the port number of the corresponding next-hop RB to the distribution tree forwarding entry.

The calculation subunit 521 can also save the port list calculated by the SPF algorithm as the port number on the root RB to the storage unit 61 as a distribution tree forwarding entry, and can send messages to multiple destination RBs at the same time through the port number. That is, the other RB obtains the port number of the corresponding next-hop RB, and adds the port number of the corresponding next-hop RB to the distribution tree forwarding entry established by itself.

Compared with the need to decapsulate and encapsulate the message in the first edge RB and the second edge RB in the second prior art, the above message sending process uses the RB identification sent by the second edge RB of the second DC to the receiving unit 51 The information can distinguish the RBs in the first DC and the RBs in the second DC, so that each RB in the first DC and the RBs in the second DC have unique identification information. The calculation subunit 521 establishes a distribution tree forwarding entry according to the RB identification information received by the receiving unit 51 in the second DC, and the sending subunit 522 sends the RB identification information received by the receiving unit 51 to other RBs in the first DC, So that the calculation subunit 521 of other RBs can establish the distribution tree forwarding table item according to the received RB identification information in the second DC, the premise that there is no need to decapsulate and encapsulate the message at the first edge RB and the second edge RB Next, the message is successfully sent to the RB of another DC, so that the RB checks the "M" field in the TRILL message header, and multicasts the message according to the distribution tree entry, reducing the use of edge RBs to support The allocation of hardware resources for packet encapsulation and decapsulation, and high packet forwarding efficiency. In addition, compared with prior art 1 where the first edge RB or the second edge RB manages all the routing tables in the first DC and the second DC, the calculation subunit 521 of the first edge RB manages the routing table of the first DC, Therefore, large-scale interconnection between DCs can be supported.

Further, as shown in FIG. 7, the receiving unit 51 is further configured to receive extended information sent by the second edge RB, where the extended information is used to describe the RB identification information and the root RB in the second DC. Correspondence between VLAN IDs.

When the second edge RB sends the RB identification information to the receiving unit 51 through BGP, in addition to sending the same content type information as in the prior art, it also sends the second additional information shown in Table 3, the second additional The information is used to indicate whether it is the root RB, the RB identification information of the root RB, and the corresponding relationship between the root RB and the VLAN. Wherein, the RB identification information of the root RB and the corresponding relationship between the root RB and the VLAN are extended information.

The processing unit 52 is further configured to establish a corresponding relationship between the RB identification information of the root RB in the first DC and the RB identification information of the root RB in the second DC according to the VLAN ID.

After the receiving unit 51 receives the second additional information sent by the second edge RB of the second DC, the processing unit 52 searches for the RB identification information of the RB corresponding to the VLANID in the first DC according to the extended information therein, that is, the VLANID, and The RB identification information, the RB identification information of the root RB in the second additional information, and the VLAN ID are stored in the storage unit 61 as a distribution tree forwarding entry.

The encapsulation unit 71 changes the egress RB name in the TRILL message header to the RB identification information of the root RB in the first DC corresponding to the original egress RB name, VLANID, established by the processing unit 52, so that the RBs in the first DC can 71. The encapsulated TRILL packet header may be forwarded to the RBs in the first DC.

As a further technical solution, compared with the decapsulation and encapsulation of the message by the first edge RB of the first DC and the second edge RB of the second DC in the prior art 2, the sending process of the above message only needs to go through the first The encapsulation unit 71 of the edge RB of a DC encapsulates the message header, and at the same time, the processing unit 52 of other RBs other than the first edge RB of the first DC does not need to perform next-hop distribution tree forwarding table for RBs in other DCs The calculation of the item reduces the allocation of hardware resources used by edge RBs to support message encapsulation and decapsulation, and the message forwarding efficiency is high, further reducing the workload of other RBs. In addition, compared with prior art 1 where the first edge RB or the second edge RB manages all the routing tables in the first DC and the second DC, the calculation subunit 521 of the first edge RB manages the routing table of the first DC, Therefore, large-scale interconnection between DCs can be supported.

Further, the calculation subunit 521 may also randomly establish distribution tree forwarding entries according to the RB identification information received by the receiving unit 51 . The random establishment is that the first edge RB takes the received RB identification information in the second DC and any tree root RB identification information in the first DC as a set of correspondences. Since any RB in the first DC can send a message to the VLAN corresponding to the VLANID through a certain port of the RB, the technical effect of interconnection between different DCs can also be achieved through the above method.

Embodiment five

An embodiment of the present invention provides a TRILL network interconnection device. The device is a second edge RB in a second data center DC. As shown in FIG. 8 , the second edge RB includes:

A sending unit 81, configured to send RB identification information to a first edge RB of the first DC.

Wherein, the RB identity information carries a combination of the RB identity ID of the root RB in the second DC and the second DC identity ID.

The RB identification information sent by the sending unit 81 to the first edge RB of the first DC can distinguish the RBs in the first DC and the RBs in the second DC, so that each RB in the first DC and the RBs in the second DC have unique identification information.

The sending unit 81 is further configured to send the RB identification information to the first edge RB through a border gateway protocol BGP.

When the sending unit 81 sends the RB identification information to the first edge RB through BGP, in addition to sending the same information as the BGP content type in the prior art, it also sends the first additional information shown in Table 2 to indicate In the additional information, whether the RB corresponding to the RB identification information is the root RB of the distribution tree, the first additional information also includes the RB identification information. Optionally, the VLAN ID corresponding to the RB identification information is also included in the first additional information, for matching the data format in the first DC. So that after the sending unit 81 of the second edge RB of the second DC sends the first additional information to the receiving unit 51 of the first edge RB of the first DC, the calculation subunit 521 of the first edge RB as the receiving end receives The RB identification information in the second DC received by the unit 51 establishes a distribution tree forwarding entry, and the sending subunit 522 sends the RB identification information received by the receiving unit 51 to other RBs in the first DC, so that the computing subunits of other RBs Unit 521 establishes a distribution tree forwarding table entry according to the received RB identification information in the second DC, which can successfully forward the message without decapsulating and encapsulating the message between the first edge RB and the second edge RB. Send it to the RB of another DC, so that the RB can check the "M" field in the TRILL message header, multicast the message according to the distribution tree entry, and reduce the edge RB's support for message encapsulation and decapsulation. The allocation of hardware resources and high packet forwarding efficiency.

Further, the sending unit 81 is further configured to send extended information to the first edge RB, where the extended information is used to describe the correspondence between the RB identification information of the root RB in the second DC and the VLAN ID.

When the sending unit 81 sends the RB identification information to the first edge RB through BGP, in addition to sending the same content type information as in the prior art, it also sends the second additional information shown in Table 3, the second additional The information is used to indicate whether it is the root RB, the RB identification information of the root RB, and the corresponding relationship between the root RB and the VLAN. Wherein, the RB identification information of the root RB and the corresponding relationship between the root RB and the VLAN are extended information. So that when the sending unit 81 of the second edge RB of the second DC sends the first additional information to the receiving unit 51 of the first edge RB of the first DC, the first edge RB as the receiving end only needs to pass the The encapsulation unit 71 of the edge RB encapsulates the message header, and at the same time, the processing unit 52 of other RBs other than the first edge RB of the first DC does not need to calculate the next-hop forwarding distribution tree forwarding entry for the RBs in other DCs , reducing the allocation of hardware resources used by edge RBs to support packet encapsulation and decapsulation, high packet forwarding efficiency, and further reducing the workload of other RBs.

Embodiment six

An embodiment of the present invention provides a TRILL network interconnection system. As shown in FIG. 9 , the system is composed of a first edge RB91 of a first DC and a second edge RB92 of a second DC.

The second edge RB92 sends RB identification information to the first edge RB91.

Wherein, the RB identity information carries a combination of the RB identity ID of the root RB in the second DC and the second DC identity ID.

The first edge RB91 receives the RB identification information sent by the second edge RB92 of the second DC.

Specifically, the second edge RB92 may send the RB identification information to the first edge RB91 through a border gateway protocol BGP.

When the second edge RB92 sends the RB identification information to the first edge RB91 through BGP, in addition to sending the same information as the BGP content type in the prior art, it also sends the first additional information shown in Table 2, so The first additional information is used to indicate whether the RB corresponding to the RB identification information in the additional information is the root RB of the distribution tree, and the first additional information also includes the RB identification information. Optionally, the VLAN ID corresponding to the RB identification information is also included in the first additional information, for matching the data format in the first DC.

The RB identification information received by the first edge RB91 establishes a distribution tree forwarding entry.

In addition to establishing the distribution tree forwarding entry for the first DC according to the RB identification information in the first DC, the first edge RB91 also establishes a distribution tree forwarding entry for each tree in the second DC according to the received RB identification information in the second DC. The distribution tree forwarding entry of the root RB.

The first edge RB91 establishes the distribution tree forwarding entry through the shortest path SPF algorithm according to the received RB identification information.

The SPF algorithm uses a certain RB as the root (ROOT) to calculate the distance to each destination RB. When calculating, each RB will calculate the topology structure diagram of the routing domain according to a unified database. The structure diagram is similar to a tree. In the SPF algorithm, it is called the shortest path tree. The first edge RB91 can obtain a path with the least number of hops between the root RB of the shortest path tree and the destination RB through the operation of the SPF algorithm, and use the RB identification information closest to the root RB in the path as the distribution tree Forward table entries to save. In addition, the first edge RB91 can also save the port list composed of the port numbers on the root RB calculated by the SPF algorithm as a distribution tree forwarding entry, so that packets can be sent to multiple destination RBs at the same time through the port numbers. That is, the first edge RB91 obtains the port number of the corresponding next-hop RB, and adds the port number of the corresponding next-hop RB into the distribution tree forwarding entry.

The first edge RB91 forwards the received RB identification information to other RBs in the first DC.

In order to enable all RBs in the first DC to send packets to RBs in the second DC, the first edge RB91 sends the received RB identification information to other RBs in the first DC. The other RBs can obtain the RB identification information of the next-hop RB used to send the message to the RB in the second DC, and save the RB identification information.

The first edge RB91 is further configured to obtain the port number of the corresponding next-hop RB, and add the port number of the corresponding next-hop RB to the distribution tree forwarding entry.

The first edge RB91 can also save the port list calculated by the SPF algorithm as the port number on the root RB as a distribution tree forwarding entry, and can send messages to multiple destination RBs at the same time through the port number, that is, the above-mentioned Obtain the port number corresponding to the next-hop RB, and add the port number corresponding to the next-hop RB to the distribution tree forwarding entry established by itself.

Compared with the need to decapsulate and encapsulate the message at the first edge RB91 and the second edge RB92 in the second prior art, the above message sending process passes through the second edge RB92 of the second DC to the first edge of the first DC The RB identification information sent by the RB91 can distinguish the RBs in the first DC and the RBs in the second DC, so that each RB in the first DC and the RBs in the second DC have unique identification information. The first edge RB91 establishes a distribution tree forwarding entry according to the received RB identification information in the second DC, and sends the received RB identification information to other RBs in the first DC, so that other RBs can receive The distribution tree forwarding entry can be established based on the RB identification information, which can successfully send the message to the RB of another DC on the premise that the first edge RB91 and the second edge RB92 do not need to decapsulate and encapsulate the message, so that The RB checks the "M" field in the TRILL message header, multicasts the message according to the distribution tree entry, reduces the allocation of hardware resources used by the edge RB to support message encapsulation and decapsulation, and improves the message forwarding efficiency. high. In addition, compared with the first edge RB91 or the second edge RB92 in the prior art to manage all the routing tables in the first DC and the second DC, the first edge RB91 manages the routing table of the first DC, so it can support large-scale interconnection between DCs.

Further, the first edge RB91 is further configured to receive the extended information sent by the second edge RB92, the extended information is used to describe the correspondence between the RB identification information of the root RB in the second DC and the VLANID.

When the second edge RB92 sends the RB identification information to the first edge RB91 through BGP, in addition to sending the same content type information as in the prior art, it also sends the second additional information shown in Table 3, the second additional information It is used to indicate whether it is the root RB, the RB identification information of the root RB, and the corresponding relationship between the root RB and the VLAN. Wherein, the RB identification information of the root RB and the corresponding relationship between the root RB and the VLAN are extended information.

The first edge RB91 is further configured to establish a correspondence between the RB identification information of the root RB in the first DC and the RB identification information of the root RB in the second DC according to the VLANID.

After receiving the second additional information sent by the second edge RB92 of the second DC, the first edge RB91 searches for the root RB identification information corresponding to the VLANID in the first DC according to the extended information in it, that is, the VLANID, and sends The RB identification information, the RB identification information of the root RB and the VLAN ID in the second additional information are saved as a distribution tree forwarding entry.

The first edge RB91 changes the egress RB name in the TRILL packet header to the RB identification information of the root RB in the first DC corresponding to the original egress RB name, VLANID, so that the TRILL encapsulated by the first edge RB91 in the first DC The packet header can be forwarded to the RBs in the first DC.

As a further technical solution, compared with the decapsulation and encapsulation of the message by the first edge RB91 of the first DC and the second edge RB92 of the second DC in the second prior art, the sending process of the above message only needs to pass through the second The edge RB of a DC encapsulates the message header, and at the same time, other RBs other than the first edge RB91 of the first DC do not need to calculate the next-hop distribution tree forwarding table items for RBs in other DCs, reducing the number of edge RBs The allocation of hardware resources used to support message encapsulation and decapsulation, the message forwarding efficiency is high, and the workload of other RBs is further reduced. In addition, compared with the first edge RB91 or the second edge RB92 in the prior art to manage all the routing tables in the first DC and the second DC, the first edge RB91 manages the routing table of the first DC, so it can support large-scale interconnection between DCs.

In the TRILL network interconnection system provided by the embodiment of the present invention, the second edge RB92, such as the second edge RB92 of the second DC, sends the RB identification information to the first edge RB91, such as the first edge RB91 of the first DC, unidirectionally. Technical solution, when the first edge RB91 of the first DC sends RB identification information to the second edge RB92 of the second DC, and the first edge RB91 of the first DC sends RB identification information to other edge RBs of other DCs, it can be realized Network interconnection among multiple DCs.

Embodiment seven

An embodiment of the present invention provides a TRILL network interconnection device. The device is a first edge RB in a first data center DC. As shown in FIG. 10 , the first edge RB includes:

The receiver 1001 is configured to receive RB identification information sent by the second edge RB of the second DC.

Wherein, the RB identity information carries a combination of the RB identity ID of the root RB in the second DC and the second DC identity ID, so that the processor 1002 can establish a distribution tree according to the RB identity information. entry, the transmitter 1003 sends the message according to the distribution tree forwarding entry.

Specifically, the receiver 1001 is further configured to receive the RB identification information sent by the second edge RB through the Border Gateway Protocol BGP.

When receiving the RB identification information, the receiver 1001 receives first additional information in addition to receiving the same type of information as the BGP content type in the prior art. The first additional information is used to indicate whether the RB corresponding to the RB identification information in the additional information is the root RB of the distribution tree, and the first additional information also includes the RB identification information. Optionally, the VLAN ID corresponding to the RB identification information is also included in the first additional information, for matching the data format in the first DC.

The processor 1002 is configured to establish a distribution tree forwarding entry according to the RB identification information received by the receiver 1001 .

In addition to establishing the distribution tree forwarding entry for the first DC according to the RB identification information in the first DC, the processor 1002 also establishes a distribution tree forwarding entry for the second DC according to the RB identification information received by the receiver 1001 in the second DC. Distribution tree forwarding entries of each tree root RB.

The processor 1002 is further configured to establish the distribution tree forwarding entry through the shortest path SPF algorithm according to the RB identification information received by the receiver 1001 .

The SPF algorithm uses a certain RB as the root (ROOT) to calculate the distance to each destination RB. When calculating, each RB will calculate the topology structure diagram of the routing domain according to a unified database. The structure diagram is similar to a tree. In the SPF algorithm, it is called the shortest path tree. The processor 1002 can obtain a path with the least number of hops between the root RB and the destination RB of the shortest path tree through the operation of the SPF algorithm, and use the RB identification information closest to the root RB in the path as the distribution tree transfer Posted items are saved to memory 1004 . In addition, the memory 1004 can store the port list composed of the port numbers on the root RB calculated by the processor 1002 according to the SPF algorithm as a distribution tree forwarding entry, so that the transmitter 1003 can send multiple destination RBs at the same time through the port numbers Send message. That is, the first edge RB obtains the port number of the corresponding next-hop RB, and adds the port number of the corresponding next-hop RB to the distribution tree forwarding entry.

The transmitter 1003 is configured to forward the RB identification information received by the receiver 1001 to other RBs in the first DC.

In order to enable all RBs in the first DC to send packets to RBs in the second DC, the transmitter 1003 sends the RB identification information received by the receiver 1001 to other RBs in the first DC. The other RBs can obtain the RB identification information of the next-hop RB used to send a message to the RB in the second DC through the processor 1002 therein, and store the RB identification information in the memory 1004 .

The processor 1002 is further configured to acquire a port number corresponding to the next-hop RB, and add the port number corresponding to the next-hop RB to the distribution tree forwarding entry.

The memory 1004 stores the port list calculated by the processor 1002 according to the SPF algorithm as the port number on the root RB as the distribution tree forwarding entry. The transmitter 1003 can send messages to multiple destination RBs at the same time through the port number, that is, the other RBs obtain the port number of the corresponding next-hop RB, and add the port number of the corresponding next-hop RB to the distribution tree established by itself In the forwarding entry.

Compared with the decapsulation and encapsulation of the message by the first edge RB of the first DC and the second edge RB of the second DC in prior art 2, the above message sending process passes the second edge RB of the second DC to the receiver The RB identification information sent by the device 1001 can distinguish the RBs in the first DC and the RBs in the second DC, so that the RBs in the first DC and the RBs in the second DC have unique identification information. The processor 1002 establishes a distribution tree forwarding entry according to the RB identification information received by the receiver 1001 in the second DC, and the transmitter 1003 sends the RB identification information received by the receiver 1001 to other RBs in the first DC, so that other The processor 1002 of the RB establishes a distribution tree forwarding table item according to the received RB identification information in the second DC, which can implement the premise that the first edge RB and the second edge RB do not need to decapsulate and encapsulate the message. The message is successfully sent to the RB of another DC, so that the RB can check the "M" field in the TRILL message header, and multicast the message according to the distribution tree entry, reducing the number of edge RBs used to support message encapsulation and the allocated amount of hardware resources for decapsulation, and the packet forwarding efficiency is high. In addition, compared with prior art 1 where the first edge RB or the second edge RB manages all the routing tables in the first DC and the second DC, the processor 1002 of the first edge RB manages the routing table of the first DC, so It can support large-scale interconnection between DCs.

Further, the receiver 1001 is further configured to receive extended information sent by the second edge RB, where the extended information is used to describe the correspondence between the RB identification information of the root RB in the second DC and the VLANID .

When the second edge RB sends the RB identification information to the receiver 1001 through BGP, in addition to sending the same content type information as in the prior art, it also sends the second additional information shown in Table 3, the second additional The information is used to indicate whether it is the root RB, the RB identification information of the root RB, and the corresponding relationship between the root RB and the VLAN. Wherein, the RB identification information of the root RB and the corresponding relationship between the root RB and the VLAN are extended information.

The processor 1002 is further configured to establish a corresponding relationship between the RB identification information of the root RB in the first DC and the RB identification information of the root RB in the second DC according to the VLANID.

After the receiver 1001 receives the second additional information sent by the second edge RB of the second DC, the processor 1002 searches for the RB identification information of the RB corresponding to the VLANID in the first DC according to the extended information therein, that is, the VLANID, and The RB identification information, the RB identification information of the root RB in the second additional information, and the VLAN ID are stored in the memory 1004 as a distribution tree forwarding entry.

The processor 1002 changes the name of the egress RB in the header of the TRILL message to the RB identification information of the root RB in the first DC corresponding to the original egress RB name, i.e. VLANID, so that the RBs in the first DC are based on the TRILL information encapsulated by the processor 1002. The packet header can be forwarded to the RBs in the first DC.

As a further technical solution, compared with the decapsulation and encapsulation of the message by the first edge RB of the first DC and the second edge RB of the second DC in the prior art 2, the sending process of the above message only needs to go through the first The processor 1002 of the edge RB of a DC encapsulates the message header, and at the same time, the processors 1002 of other RBs other than the first edge RB of the first DC do not need to perform next-hop distribution tree forwarding table for RBs in other DCs The calculation of the item reduces the allocation of hardware resources used by edge RBs to support message encapsulation and decapsulation, and the message forwarding efficiency is high, further reducing the workload of other RBs. In addition, compared with prior art 1 where the first edge RB or the second edge RB manages all the routing tables in the first DC and the second DC, the processor 1002 of the first edge RB manages the routing table of the first DC, so It can support large-scale interconnection between DCs.

Further, the processor 1002 may also randomly establish distribution tree forwarding entries according to the RB identification information received by the receiver 1001 . The random establishment is that the first edge RB takes the received RB identification information in the second DC and any tree root RB identification information in the first DC as a set of correspondences. Since any RB in the first DC can send a message to the VLAN corresponding to the VLANID through a certain port of the RB, the technical effect of interconnection between different DCs can also be achieved through the above method.

Embodiment eight

An embodiment of the present invention provides a TRILL network interconnection device. The device is a second edge RB in a second data center DC. As shown in FIG. 11 , the second edge RB includes:

The transmitter 1101 is configured to send RB identification information to a first edge RB of the first DC.

Wherein, the RB identity information carries a combination of the RB identity ID of the root RB in the second DC and the second DC identity ID.

The RB identification information sent by the transmitter 1101 to the first edge RB of the first DC can distinguish the RBs in the first DC and the RBs in the second DC, so that each RB in the first DC and the RBs in the second DC have unique identification information.

The transmitter 1101 is further configured to send the RB identification information to the first edge RB through a border gateway protocol BGP.

When the transmitter 1101 sends the RB identification information to the first edge RB through BGP, in addition to sending the same information as the BGP content type in the prior art, it also sends the first additional information shown in Table 2, the The first additional information is used to indicate whether the RB corresponding to the RB identification information in the additional information is the root RB of the distribution tree, and the first additional information also includes the RB identification information. Optionally, the VLAN ID corresponding to the RB identification information is also included in the first additional information, for matching the data format in the first DC. So that when the transmitter 1101 of the second edge RB of the second DC sends the first additional information to the receiver 1001 of the first edge RB of the first DC, the processor 1002 of the first edge RB as the receiving end according to the receiver 1001 1001 The received RB identification information in the second DC establishes a distribution tree forwarding entry, and the transmitter 1003 sends the RB identification information received by the receiver 1001 to other RBs in the first DC, so that the processors 1002 of other RBs according to The received RB identification information in the second DC establishes a distribution tree forwarding table entry, which can successfully send the message to another edge RB and the second edge RB without decapsulating and encapsulating the message. In the RB of a DC, the RB checks the "M" field in the TRILL message header, multicasts the message according to the distribution tree entry, and reduces the hardware resources used by the edge RB to support message encapsulation and decapsulation The allocation amount is high, and the message forwarding efficiency is high.

Further, the transmitter 1101 is further configured to send extended information to the first edge RB, where the extended information is used to describe the correspondence between the RB identification information of the root RB in the second DC and the VLAN ID.

When the transmitter 1101 sends the RB identification information to the first edge RB through BGP, in addition to sending the same content type information as in the prior art, it also sends the second additional information shown in Table 3, the second additional The information is used to indicate whether it is the root RB, the RB identification information of the root RB, and the corresponding relationship between the root RB and the VLAN. Wherein, the RB identification information of the root RB and the corresponding relationship between the root RB and the VLAN are extended information. So that when the transmitter 1101 of the second edge RB of the second DC sends the second additional information to the receiver 1001 of the first edge RB of the first DC, the first edge RB as the receiving end only needs to pass the The processor 1002 of the edge RB encapsulates the message header, and at the same time, the processors 1002 of other RBs other than the first edge RB of the first DC do not need to calculate the next-hop distribution tree forwarding table items for the RBs in other DCs , reducing the allocation of hardware resources used by edge RBs to support packet encapsulation and decapsulation, high packet forwarding efficiency, and further reducing the workload of other RBs.

Embodiment nine

An embodiment of the present invention provides a TRILL network interconnection system. As shown in FIG. 12 , the system consists of a first edge RB1201 and the second edge RB1202.

The second edge RB1202, for example, the second edge RB1202 of the second DC, sends RB identification information to the first edge RB1201, for example, the first edge RB1201 of the first DC.

Wherein, the RB identity information carries a combination of the RB identity ID of the root RB in the second DC and the second DC identity ID.

The first edge RB1201 receives the RB identification information sent by the second edge RB1202 of the second DC.

Specifically, the second edge RB1202 may send the RB identification information to the first edge RB1201 through a border gateway protocol BGP.

When the second edge RB1202 sends the RB identification information to the first edge RB1201 through BGP, in addition to sending the same information as the BGP content type in the prior art, it also sends the first additional information shown in Table 2, so The first additional information is used to indicate whether the RB corresponding to the RB identification information in the additional information is the root RB of the distribution tree, and the first additional information also includes the RB identification information. Optionally, the VLAN ID corresponding to the RB identification information is also included in the first additional information, for matching the data format in the first DC.

The RB identification information received by the first edge RB1201 establishes a distribution tree forwarding entry.

In addition to establishing the distribution tree forwarding entry for the first DC according to the RB identification information in the first DC, the first edge RB 1201 also establishes a distribution tree forwarding entry for each tree in the second DC according to the received RB identification information in the second DC. The distribution tree forwarding entry of the root RB.

The first edge RB1201 establishes the distribution tree forwarding entry through the shortest path SPF algorithm according to the received RB identification information.

The SPF algorithm uses a certain RB as the root (ROOT) to calculate the distance to each destination RB. When calculating, each RB will calculate the topology structure diagram of the routing domain according to a unified database. The structure diagram is similar to a tree. In the SPF algorithm, it is called the shortest path tree. The first edge RB1201 can obtain a path with the least number of hops between the root RB of the shortest path tree and the destination RB through the operation of the SPF algorithm, and use the RB identification information closest to the root RB in the path as the distribution tree Forward table entries to save. In addition, the first edge RB1201 can also save the port list composed of port numbers on the root RB calculated by the SPF algorithm as a distribution tree forwarding entry, so that packets can be sent to multiple destination RBs at the same time through the port numbers. That is, the first edge RB1201 obtains the port number of the corresponding next-hop RB, and adds the port number of the corresponding next-hop RB into the distribution tree forwarding entry.

The first edge RB1201 forwards the received RB identification information to other RBs in the first DC.

In order to enable all RBs in the first DC to send packets to RBs in the second DC, the first edge RB 1201 sends the received RB identification information to other RBs in the first DC. The other RBs can obtain the RB identification information of the next-hop RB used to send the message to the RB in the second DC, and save the RB identification information.

The first edge RB1201 is further configured to acquire the port number of the corresponding next-hop RB, and add the port number of the corresponding next-hop RB to the distribution tree forwarding entry.

The first edge RB1201 can also save the port list calculated by the SPF algorithm as the port number on the root RB as a distribution tree forwarding entry, and can send messages to multiple destination RBs at the same time through the port number, that is, The other RB obtains the port number of the corresponding next-hop RB, and adds the port number of the corresponding next-hop RB to the distribution tree forwarding entry established by itself.

Compared with the decapsulation and encapsulation of the message by the first edge RB1201 of the first DC and the second edge RB1202 of the second DC in the second prior art, the above message sending process passes through the second edge RB1202 of the second DC to the second The RB identification information sent by the first edge RB 1201 of a DC can distinguish the RBs in the first DC and the RBs in the second DC, so that each RB in the first DC and in the second DC has unique identification information. The first edge RB1201 establishes a distribution tree forwarding entry according to the received RB identification information in the second DC, and sends the received RB identification information to other RBs in the first DC, so that other RBs receive The distribution tree forwarding table entry can be established based on the RB identification information in the first edge RB1201 and the second edge RB1202, and the message can be successfully sent to the RB of another DC without decapsulating and encapsulating the message. In order for the RB to check the "M" field in the TRILL message header, multicast the message according to the distribution tree entry, reduce the allocation of hardware resources used by the edge RB to support message encapsulation and decapsulation, and message forwarding efficient. In addition, compared with the first edge RB1201 or the second edge RB1202 in the prior art to manage all the routing tables in the first DC and the second DC, the first edge RB1201 manages the routing table of the first DC, so it can support large-scale interconnection between DCs.

Further, the first edge RB1201 is further configured to receive the extended information sent by the second edge RB1202, where the extended information is used to describe the correspondence between the RB identification information of the root RB in the second DC and the VLANID.

When the second edge RB1202 sends the RB identification information to the first edge RB1201 through BGP, in addition to sending the same content type information as in the prior art, it also sends the second additional information shown in Table 3, the second additional information It is used to indicate whether it is the root RB, the RB identification information of the root RB, and the corresponding relationship between the root RB and the VLAN. Wherein, the RB identification information of the root RB and the corresponding relationship between the root RB and the VLAN are extended information.

The first edge RB 1201 is further configured to establish a corresponding relationship between the RB identification information of the root RB in the first DC and the RB identification information of the root RB in the second DC according to the VLAN ID.

After receiving the second additional information sent by the second edge RB1202 of the second DC, the first edge RB1201 searches for the RB identification information of the RB corresponding to the VLANID in the first DC according to the extended information in it, that is, the VLANID, and sends the The RB identification information and the RB identification information of the root RB in the second additional information and the VLAN ID are saved as a distribution tree forwarding entry.

The first edge RB1201 changes the egress RB name in the TRILL packet header to the RB identification information of the root RB in the first DC corresponding to the original egress RB name, VLANID, so that the TRILL encapsulated by the first edge RB1201 in the first DC The packet header can be forwarded to the RBs in the first DC.

As a further technical solution, compared with the decapsulation and encapsulation of the message by the first edge RB1201 of the first DC and the second edge RB1202 of the second DC in the second prior art, the sending process of the above message only needs to pass through the second The edge RB of a DC encapsulates the packet header, and at the same time, other RBs other than the first edge RB1201 of the first DC do not need to calculate the next-hop distribution tree forwarding table items for RBs in other DCs, reducing the number of edge RBs The allocation of hardware resources used to support message encapsulation and decapsulation, the message forwarding efficiency is high, and the workload of other RBs is further reduced. In addition, compared with the first edge RB1201 or the second edge RB1202 in the prior art to manage all the routing tables in the first DC and the second DC, the first edge RB1201 manages the routing table of the first DC, so it can support large-scale interconnection between DCs.

In the TRILL network interconnection system provided by the embodiment of the present invention, the second edge RB1202, such as the second edge RB1202 of the second DC, sends RB identification information to the first edge RB1201, such as the first edge RB1201 of the first DC. Technical solution, when the first edge RB1201 of the first DC sends RB identification information to the second edge RB1202 of the second DC, and the first edge RB1201 of the first DC sends RB identification information to other edge RBs of other DCs, it can be realized Network interconnection among multiple DCs.

Those skilled in the art can clearly understand that for the convenience and brevity of the description, only the division of the above-mentioned functional modules is used as an example for illustration. In practical applications, the above-mentioned function allocation can be completed by different functional modules according to needs. The internal structure of the device is divided into different functional modules to complete all or part of the functions described above. For the specific working process of the above-described system, device, and unit, reference may be made to the corresponding process in the foregoing method embodiments, and details are not repeated here.

In the several embodiments provided in this application, it should be understood that the disclosed system, device and method can be implemented in other ways. For example, the device embodiments described above are only illustrative. For example, the division of the modules or units is only a logical function division, and there may be other division methods in actual implementation. For example, multiple units or components can be Incorporation may either be integrated into another system, or some features may be omitted, or not implemented. In another point, the mutual coupling or direct coupling or communication connection shown or discussed may be through some interfaces, and the indirect coupling or communication connection of devices or units may be in electrical, mechanical or other forms.

The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place, or may be distributed to multiple network units. Part or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, each unit may exist separately physically, or two or more units may be integrated into one unit. The above-mentioned integrated units can be implemented in the form of hardware or in the form of software functional units.

If the integrated unit is realized in the form of a software function unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the essence of the technical solution of the present invention or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium , including several instructions for causing a computer device (which may be a personal computer, server, or network device, etc.) or a processor (processor) to execute all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-OnlyMemory), random access memory (RAM, RandomAccessMemory), magnetic disk or optical disk and other media that can store program codes.

The above is only a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Anyone skilled in the art can easily think of changes or substitutions within the technical scope disclosed in the present invention. Should be covered within the protection scope of the present invention. Therefore, the protection scope of the present invention should be determined by the protection scope of the claims.

Claims (19)

1. the translucent interconnected TRILL network interconnected method of multi-link, it is characterized in that, described method comprises:

The first edge routing bridge RB of the first data center DC receives the RB identification information of the second edge RB transmission of the 2nd DC, carries the RB identify label ID of tree root RB in described 2nd DC and the combination of described 2nd DC identify label ID in described RB identification information;

Described first edge RB sets up distribution tree forwarding-table item according to described RB identification information, to send message according to described distribution tree forwarding-table item; Described first edge RB and each RB in a described DC sets up the distribution tree forwarding-table item for a described DC inside according to RB identification information in a described DC, also sets up the distribution tree forwarding-table item for each tree root RB in described 2nd DC according to the RB identification information in described 2nd DC received.

2. method according to claim 1, is characterized in that, the first edge RB of described first data center DC receives the RB identification information of the second edge RB transmission of the 2nd DC, specifically comprises:

Described first edge RB receives the described RB identification information that described second edge RB is sent by Border Gateway Protocol (BGP).

3. method according to claim 2, is characterized in that, described first edge RB sets up distribution tree forwarding-table item according to described RB identification information, specifically comprises:

Described first edge RB sets up described distribution tree forwarding-table item according to described RB identification information by shortest-path first algorithm SPF algorithm;

Described RB identification information is transmitted to other RB in a described DC by described first edge RB, so that other RB described set up the distribution tree forwarding-table item of self by described SPF algorithm according to described RB identification information.

4. method according to claim 3, is characterized in that, described first edge RB sets up described distribution tree forwarding-table item according to described RB identification information by shortest path SPF algorithm, specifically comprises:

Described first edge RB obtains the port numbers of corresponding down hop RB, the port numbers of described corresponding down hop RB is added in described distribution tree forwarding-table item;

Other RB described set up self distribution tree forwarding-table item by described SPF algorithm according to described RB identification information, specifically comprise:

Other RB described obtain the port numbers of corresponding down hop RB, the port numbers of described corresponding down hop RB are added in the distribution tree forwarding-table item self set up.

5. method according to claim 1, is characterized in that, the first edge RB of described first data center DC receives in the RB identification information of the second edge RB transmission of the 2nd DC and also carries:

Virtual LAN VLAN identify label ID, described VLANID are for identifying the VLAN that in a described DC, in tree root RB and described 2nd DC, tree root RB belongs to together;

The first edge RB of described first data center DC receives the RB identification information of the second edge RB transmission of the 2nd DC, specifically comprises:

Described first edge RB receives the extend information that described second edge RB sends, and described extend information is for describing the corresponding relation between the RB identification information of tree root RB in described 2nd DC and VLANID.

6. method according to claim 5, is characterized in that, described first edge RB sets up distribution tree forwarding-table item according to described RB identification information, specifically comprises:

Described first edge RB sets up the corresponding relation of the RB identification information of tree root RB in the RB identification information of tree root RB in a described DC and described 2nd DC according to VLANID, thus described extend information is converted into local distribution tree forwarding-table item.

7. the translucent interconnected TRILL network interconnected method of multi-link, it is characterized in that, described method comprises:

The second edge RB of the second data center DC sends RB identification information to the first edge routing bridge RB of a DC, the RB identify label ID of tree root RB in described 2nd DC and the combination of described 2nd DC identify label ID is carried in described RB identification information, so that described first edge RB sets up distribution tree forwarding-table item according to described RB identification information, send message according to described distribution tree forwarding-table item; Described first edge RB and each RB in a described DC sets up the distribution tree forwarding-table item for a described DC inside according to RB identification information in a described DC, also sets up the distribution tree forwarding-table item for each tree root RB in described 2nd DC according to the RB identification information in described 2nd DC received.

8. method according to claim 7, is characterized in that, the second edge RB of described second data center DC, to the first edge RB transmission RB identification information of a DC, specifically comprises:

Described second edge RB sends described RB identification information by Border Gateway Protocol (BGP) to described first edge RB.

9. method according to claim 8, is characterized in that, described second edge RB also carries in the described RB identification information of described first edge RB transmission:

Virtual LAN VLAN identify label ID, described VLANID are for identifying the VLAN that in a described DC, in tree root RB and described 2nd DC, tree root RB belongs to together;

The second edge RB of described second data center DC, to the first edge RB transmission RB identification information of a DC, specifically comprises:

Described second edge RB sends extend information to described first edge RB, and described extend information is for describing the corresponding relation between the RB identification information of tree root RB in described 2nd DC and VLANID.

10. the translucent interconnected TRILL network interconnect device of multi-link, described device is the first edge routing bridge RB in the first data center DC, it is characterized in that, described first edge RB comprises:

Receiving element, the RB identification information that the second edge RB for receiving the 2nd DC sends, carries the RB identify label ID of tree root RB in described 2nd DC and the combination of described 2nd DC identify label ID in described RB identification information;

Processing unit, sets up distribution tree forwarding-table item for the described RB identification information received according to described receiving element; Described first edge RB and each RB in a described DC sets up the distribution tree forwarding-table item for a described DC inside according to RB identification information in a described DC, also sets up the distribution tree forwarding-table item for each tree root RB in described 2nd DC according to the RB identification information in described 2nd DC received.

11. devices according to claim 10, is characterized in that, described receiving element is also for receiving the described RB identification information that described second edge RB is sent by Border Gateway Protocol (BGP).

12. devices according to claim 11, is characterized in that, described processing unit specifically comprises:

Computation subunit, sets up described distribution tree forwarding-table item for the described RB identification information received according to described receiving element by shortest path SPF algorithm;

Send subelement, the described RB identification information for being received by described receiving element is transmitted to other RB in a described DC.

13. devices according to claim 12, is characterized in that, the port numbers of described corresponding down hop RB, also for obtaining the port numbers of corresponding down hop RB, is added in described distribution tree forwarding-table item by described computation subunit.

14. devices according to claim 10, it is characterized in that, described receiving element is also for receiving the extend information that described second edge RB sends, and described extend information is for describing the corresponding relation between the RB identification information of tree root RB in described 2nd DC and VLANID.

15. device according to claim 14, is characterized in that, described processing unit is also for setting up the corresponding relation of the RB identification information of tree root RB in the RB identification information of tree root RB in a described DC and described 2nd DC according to VLANID.

16. 1 kinds of translucent interconnected TRILL network interconnect devices of multi-link, described device is the second edge routing bridge RB in the second data center DC, it is characterized in that, described second edge RB comprises:

Transmitting element, sends RB identification information for the first edge RB to a DC, carries the RB identify label ID of tree root RB in described 2nd DC and the combination of described 2nd DC identify label ID in described RB identification information; Described first edge RB and each RB in a described DC sets up the distribution tree forwarding-table item for a described DC inside according to RB identification information in a described DC, also sets up the distribution tree forwarding-table item for each tree root RB in described 2nd DC according to the RB identification information in described 2nd DC received.

17. devices according to claim 16, is characterized in that, described transmitting element is also for sending described RB identification information by Border Gateway Protocol (BGP) to described first edge RB.

18. devices according to claim 17, it is characterized in that, described transmitting element is also for sending extend information to described first edge RB, and described extend information is for describing the corresponding relation between the RB identification information of tree root RB in described 2nd DC and VLANID.

19. 1 kinds of translucent interconnected TRILL network interconnected systemss of multi-link, it is characterized in that, described system is made up of the second edge RB in the first edge routing bridge RB in claim 10 to claim 15 described in any one and claim 16 to claim 18 described in any one.