WO2014079246A1

WO2014079246A1 - Multicast flow forwarding implementation method and routing bridge (rb)

Info

Publication number: WO2014079246A1
Application number: PCT/CN2013/082327
Authority: WO
Inventors: 翟洪军; 代雪会; 廖婷
Original assignee: 中兴通讯股份有限公司
Priority date: 2012-11-20
Filing date: 2013-08-27
Publication date: 2014-05-30
Also published as: CN102970231A; CN102970231B

Abstract

A multicast flow forwarding implementation method and routing bridge (RB). Each routing bridge (RB) forwarding a received uplink or downlink multicast flow according to allocation information of multicast trees, wherein multicast trees in the allocation information of the multicast trees are respectively allocated in an uplink direction and a downlink direction. When the downlink direction fails, multicast flow forwarding in the uplink direction will not be influenced.

Description

Multicast stream forwarding implementation method and routing bridge (RB)

Technical field

The present invention relates to network communication technologies, and more particularly to a multicast stream forwarding implementation method and a routing bridge (RB). Background technique

TRILL is the abbreviation of Transparent Interconnection of Lots of Links, which is used to solve the problem of multipath (or called Layer 2 Multiple Path) in the data center large Layer 2 network. L2MP is implemented by introducing IS-IS (Intermediate System to Intermediate System) routing protocol into the Layer 2 network. In a TRILL network, a device running the TRILL protocol is called RBridge (also referred to as "RB"), and a Nickname (alias) uniquely identifies an RBridge. At the entrance of the TRILL network, it is responsible for encapsulating the original data frame of the end device into the TRILL format (that is, adding the TRILL header and the outer frame header in front of the original data frame, and the TRILL header mainly includes the TRILL network entry and the egress routing bridge. The routing bridge of the Nickname and hop counts and injected into the TRILL network is called Ingress; at the exit of the TRILL network, it is responsible for decapsulating the TRILL data frame into the frame, which Ingress is imported into the TRILL network and forming the MAC information table {D-MAC , Ingress Nickname , ...... }. Because the Ingress and Egress routing bridges are at the edge of the TRILL network, they are also known as Edge RBridges.

In order to avoid loops, at the border of the TRILL network, only one RBridge can serve one end system in any VLAN (Virtual Local Area Network). This RBridge is called the service provider of this end system, for example. On the shared link, the designated forwarding device (AF) of the VLAN. Although this can effectively avoid the loop, it also causes some problems. For example, after the AF switch, the Ingress- Nickname changes in some MAC entries on the remote Egress; the end-system multi-homing through the point-to-point link (Multi -homing, than ^ port: Link Aggregation Group (LAG)) When multiple RBridges are used, load sharing on the RB link causes some MAC entries on the remote RBridge. Ingress—The Nickname fluctuates frequently (called a flip-flop), causing out-of-order traffic and loss of traffic, causing session interruptions.

To this end, the TRILL Working Group proposed the concept of a Routing Bridge Group (RBG) or Virtual RBridge (RBv). In an RBv, the group members share a Nickname, called the group Nickname. The RBs in the RBv solve the above flip-flop problem by encapsulating the data frame with the group Nickname when importing the data frame into the TRILL network.

To avoid the temporary loop in multicast packet forwarding, the TRILL protocol specifies that the RBridge must perform RPF (Reverse Path Forwarding) check when receiving the multicast frame. The RPF check requires that any one of the routing bridges in any one of the multicast trees can receive the multicast frame sent by the other RBridge along the tree from one interface. The frame must be discarded if the check fails. The working group document draft-ietf-trill-cmt-01.txt (referred to as CMT, Coordinated Multi-Destination Tree, calibrated multicast tree) focuses on and solves many problems by allocating available multicast trees among member RBs in RBv. The RPF problem that exists when a member RB in an RBv imports (upstream) a multicast data frame. For a multicast data frame that leaves the (downlink) TRILL network through the member RBs in the RBv, the CMT specifies that the RBs in any RBv can only export the traffic distributed along the multicast tree allocated to it to the TRILL network, avoiding the end system. The risk of receiving multiple multicast packets. As shown in Figure 1, assuming five multicast trees, RB1, RB2, and RB3 advertise {RBI, RBv, Tl, T4}, {RB2, RBv, Τ2, Τ5} and {RBI, RBv, T3}, respectively. 4. Let RB4 send multicast traffic along T1 (the multicast tree shown by the thick solid line in Figure 1). In this RBv, the multicast stream can only be exported to the TRILL network by RB1, and RB2 and RB3 are discarded. The multicast stream. However, this mechanism has the following problems when the RB or access link fails.

First, when an RB fails, the multicast tree is redistributed between RBs in the RBv. During this period, the receiver of the multicast stream cannot receive the expected data stream. In addition, when the multicast tree is redistributed. After that, the remote RB (such as RB4 in Figure 1) may re-select another multicast tree, which will result in longer service interruption.

Second, when the RB link of LAG1 under RB1 is normal for the traffic sent along T1 (that is, when there is no fault), the Customer Edge (CE) devices CE1 and CE2 can receive the traffic; When RB1-CE1 fails in RB link under LAG1, if another multicast tree is reselected on RB4, assuming T2, then according to the current CMT document, RB2 The multicast stream will be exported. One problem with this is that because the RB link of LAG1 fails, the receiver on LAG2 also receives traffic from another tree. That is to say, when there are multiple LAGs in one RBv, if only one LAG RB link fails, traffic to other LAGs will also be affected.

Summary of the invention

The technical problem to be solved by the present invention is to provide a multicast stream forwarding implementation method and a routing bridge (RB) to solve the problem that the multicast stream forwarding is greatly affected when the existing fault occurs.

To solve the above technical problem, the present invention provides a method for implementing multicast stream forwarding, and each routing bridge (RB) forwards the received uplink or downlink multicast stream according to the allocated information of the stored multicast tree, where the multicast The multicast tree in the allocation information of the tree is allocated in the uplink direction and the downlink direction, respectively.

To solve the above technical problem, the present invention also provides a routing bridge (RB), the routing bridge includes:

The allocation information storage unit is configured to: store allocation information of the multicast tree, where the multicast tree is allocated in an uplink direction and a downlink direction, respectively;

The multicast stream forwarding unit is configured to: forward the uplink and/or downlink multicast streams according to the allocation information of the multicast tree.

In the method and the routing bridge of the embodiment of the present invention, the multicast tree in the allocation information of the multicast tree is allocated in the uplink direction and the downlink direction respectively. Therefore, when the downlink direction fails, the uplink direction group is not affected. Broadcast forwarding, which in turn enables high-reliability access of the end system to the TRILL network.

BRIEF abstract

Figure 1 is a schematic diagram of the allocation of a CMT document to a multicast tree;

2 is a schematic flowchart of Embodiment 1 of a method for implementing multicast stream forwarding according to the present invention;

3 is a schematic flowchart of Embodiment 2 of a method for implementing multicast stream forwarding according to the present invention;

4 is a schematic flowchart of Embodiment 3 of a method for implementing multicast stream forwarding according to the present invention;

5 is a schematic diagram of a process of multicast stream forwarding in Embodiment 3 of a method for implementing multicast stream forwarding according to the present invention; 6 is a schematic flowchart of Embodiment 4 of a method for implementing multicast stream forwarding according to the present invention; FIG. 7 and FIG. 8 are schematic diagrams of allocation of a multicast tree used in Application Examples 1 to 4 of the present invention;

9-11 are block diagrams showing the structure of an embodiment of a routing bridge (RB) according to the present invention. Preferred embodiment of the invention

The technical solutions of the present invention are further described in detail below with reference to the accompanying drawings and specific embodiments to enable those skilled in the art to understand the invention. . It should be noted that the embodiments in the present application and the features in the embodiments may be combined with each other without conflict.

Example 1

As shown in FIG. 2, in the embodiment 1 of the present invention, a method for forwarding a multicast stream, each routing bridge (RB) forwards the received uplink or downlink multicast stream according to the allocation information of the multicast tree, and allocates the multicast tree. The multicast tree in the information is allocated in the uplink direction and the downlink direction respectively. Specifically, the method includes: Step 201: The routing bridge (RB) acquires and stores the allocation information of the multicast tree of the routing bridge group (RBv). The multicast tree in the allocation information of the multicast tree is allocated in an uplink direction and a downlink direction, respectively;

The allocation information of the multicast tree may be specifically generated by using a configuration manner or each routing bridge is obtained based on the same allocation algorithm. Preferably, the allocation may be performed based on a load balancing policy.

The present invention does not limit the specific method for the routing bridge (RB) to obtain the allocation information of the multicast tree of the routing bridge group (RBv). In the uplink direction, the multicast tree can be allocated according to the existing CMT mechanism.

Obviously, the mapping relationship between the routing bridge and the multicast tree is mainly established, which may be described as assigning a multicast tree to a routing bridge, or as a routing bridge to a multicast tree. There is no substantive difference.

A routing bridge to which a multicast tree is assigned is the forwarding device of the multicast tree.

Step 202: Each routing bridge (RB) performs uplink and/or downlink according to allocation information of the multicast tree. Multicast stream forwarding.

The process of performing uplink and/or downlink multicast stream forwarding by each RB according to the allocation information of the multicast tree includes:

Receiving a multicast stream in the uplink or downlink direction along the multicast tree;

Determining whether it is a forwarding device of the multicast tree in a corresponding direction according to the allocation information; if yes, forwarding the multicast stream, otherwise discarding the multicast stream.

In the above embodiment, the multicast tree is allocated in the uplink direction and the downlink direction between the routing bridges (RBs) of the routing bridge group (RBv), so that the downlink routing bridge or its corresponding chain can be made. When a road fails, it does not affect the forwarding of upstream multicast streams.

Example 2

As shown in FIG. 3, in the embodiment 2 of the present invention, a method for forwarding a multicast stream, each routing bridge (RB) forwards the received uplink or downlink multicast stream according to the allocation information of the multicast tree, and allocates the multicast tree. The multicast tree in the information is allocated in the uplink direction and the downlink direction respectively. Specifically, the method includes:

Step 301: The routing bridge (RB) acquires and stores the allocation information of the multicast tree of the routing bridge group (RBv), where the multicast tree in the allocation information of the multicast tree is respectively in the uplink direction and the downlink direction.

Area Network, LAN )

For example, if there are 6 multicast trees and 3 LAGs, 6 multicast trees are allocated in all 3 LAGs. The allocation information of the multicast tree may be specifically configured by using a configuration manner or the routing bridges are obtained by using the same allocation algorithm. Preferably, the multicast tree may be balanced according to the number of multicast trees corresponding to the multicast flows that each RB is responsible for forwarding. .

The present invention does not limit the specific method for the routing bridge (RB) to obtain the allocation information of the multicast tree of the routing bridge group (RBv).

Obviously, the mapping relationship between the routing bridge and the multicast tree is mainly established, which may be described as assigning a multicast tree to a routing bridge, or as a routing bridge to a multicast tree. There is no substantive difference. A routing bridge to which a multicast tree is assigned is the forwarding device of the multicast tree.

Step 302: Each routing bridge (RB) performs uplink and/or downlink multicast forwarding according to the allocation information of the multicast tree.

The process of forwarding the downlink multicast stream by using the multicast tree allocated by the RB in the downlink direction includes: receiving a multicast stream in a downlink direction along the multicast tree;

Determining, according to the allocation information, whether the multicast device is a forwarding device in a downlink direction of a link aggregation group (LAG) or a local local area network (LAN);

If yes, the multicast stream is forwarded, otherwise the multicast stream is discarded.

In the above embodiment, the multicast tree is allocated in the uplink direction and the downlink direction between the routing bridges (RBs) of the routing bridge group (RBv), so that the downlink routing bridge or its corresponding chain can be made. When a path fails, it does not affect the forwarding of the upstream multicast stream. In addition, the multicast tree is allocated in the downlink direction in each link aggregation group (LAG) or local area network (LAN), which can make one of the LAGs or LANs. When a link fails, it does not affect the multicast forwarding of other LAGs or LANs.

Example 3

The third embodiment of the present invention focuses on forwarding the multicast stream in the downlink direction, and each routing bridge (RB) forwards the received uplink or downlink multicast stream according to the allocation information of the multicast tree, where the multicast tree is allocated information. The multicast tree is allocated in the uplink direction and the downlink direction. Specifically, as shown in FIG. 4, the multicast stream forwarding implementation method includes:

Step 401: The routing bridge (RB) acquires and stores allocation information of the multicast tree of the routing bridge group (RBv) in the downlink direction, where the multicast tree is in each link aggregation group (LAG) or local area network. (LAN), a routing bridge (RB) allocated by a multicast tree in the downlink direction includes an active forwarding device and a standby forwarding device;

The specific method for the routing bridge (RB) to obtain the allocation information of the multicast tree of the routing bridge group (RBv) is not limited. Obviously, the mapping relationship between the routing bridge and the multicast tree is mainly established, which may be described as assigning a multicast tree to a routing bridge, or as a routing bridge to a multicast tree. There is no substantive difference.

A routing bridge to which a multicast tree is assigned is the forwarding device of the multicast tree. In this embodiment, one or all of the multicast trees in the downlink direction are assigned one primary forwarding device and one standby forwarding device. When the primary forwarding device and its corresponding link are normal, when the primary forwarding device implements the broadcast forwarding of the corresponding multicast tree, the multicast forwarding function of the standby forwarding device is enabled.

Step 402: Each routing bridge (RB) forwards the downlink multicast stream according to the allocation information of the multicast tree in the downlink direction.

As shown in FIG. 5, the process of forwarding the downlink multicast stream by each routing bridge (RB) according to the multicast tree allocated in the downlink direction includes:

Step 501: Receive a multicast stream in a downlink direction of the multicast tree.

Step 502: Determine, according to the allocation information, whether it is an active forwarding device in a downlink direction of the multicast tree.

Specifically, the routing bridge can establish a flag bit for each multicast tree allocated to it to indicate whether the current routing bridge is the primary forwarding device or the standby forwarding device of the multicast tree.

Step 503: If yes, forward the multicast stream, otherwise discard the multicast stream.

In the above embodiment, although the downlink direction is mainly described, the multicast tree may be allocated in the uplink direction and the downlink direction respectively between the routing bridges (RBvs) of the routing bridge group (RBv), so that When the downlink routing bridge or its corresponding link fails, the uplink multicast stream is not affected. In addition, the multicast tree is in the downlink aggregation group (LAG) or local area network (LAN). Allocation, when one of the LAG or LAN links fails, does not affect the multicast stream forwarding of other LAGs or LANs; the routing bridge (RB) allocated by the same multicast tree in the downlink direction includes a primary forwarding. The device and an alternate forwarding device may be configured to upgrade the backup forwarding device of the affected multicast tree to the primary forwarding device of the multicast tree to implement multicast forwarding when the primary forwarding device or its corresponding link fails. Send, thus implementing the routing bridge group (RBv) Fast protection.

Specifically, a multicast tree explicitly allocates an alternate forwarding device and a routing bridge in the downlink direction.

When (RB) or its corresponding link fails, the fault-related node sends fault information to all other routing bridges (RBs) in the RBv; the alternate forwarding device of the affected multicast tree is upgraded to the multicast tree. The primary forwarding device.

Specifically, if the RB is faulty, the CE is a fault-related node; if the RB link is faulty, and the RB and the CE corresponding to the link sense the fault, the fault-related point may be RB and/or CE.

The content of the fault information is different for different multicast tree allocation information generation methods, such as:

If the configuration mode is adopted, and each RB is configured with only the allocation information of its own (ie, the current RB), that is, the allocation information acquired and stored by each routing bridge (RB) includes only the allocation information of its own multicast tree. The fault information needs to carry the affected multicast tree information. If it is allocated in the LAG or the LAN, it also needs to carry the relevant LAG or LAN information, so that the standby forwarding device that receives the multicast fault information can determine that it needs to turn. The primary forwarding device of the affected multicast tree; if the configuration mode is adopted, and each RB is configured not only with its own allocation information but also with other RB allocation information (configuration synchronization), or each RB uses the same allocation. Algorithm (algorithm synchronization), each RB not only knows its own allocation information, but also knows the allocation information of other RBs, or each RB only configures its own allocation information, and realizes synchronization (interactive synchronization) of allocation information through interaction. In this case, the allocation information of each RB is synchronized, that is, the points of the multicast tree acquired and stored by each routing bridge (RB) in the RBv. The information includes the allocation information of the multicast tree of itself and other RBs, and the multicast fault information does not carry the affected multicast tree information and/or LAG or LAN information, and the alternate forwarding device of the affected multicast tree also It can be smoothly transferred to the primary forwarding device to perform the corresponding multicast stream forwarding.

The above fault information is used to notify all other routing bridges (RBs) in the RBv by means of multicast, and alternatively, the allocation information of the multicast trees acquired and stored by the routing bridges (RBs) of the same RBv includes itself and The allocation information of the multicast trees of the other RBs may be specifically synchronized by the foregoing interaction synchronization or configuration synchronization or algorithm; when the routing bridge (RB) or its corresponding link fails, the fault-related nodes are allocated according to the multicast tree. The information is sent to the alternate forwarding device of the affected multicast tree. Information; The alternate forwarding device that receives the fault information is upgraded to the primary forwarding device of the multicast tree.

Example 4

Each routing bridge (RB) forwards the received uplink or downlink multicast stream according to the allocation information of the multicast tree, and the multicast tree in the allocation information of the multicast tree is allocated in the uplink direction and the downlink direction respectively. The embodiment of the present invention focuses on the forwarding of the multicast stream in the downlink direction. Specifically, as shown in FIG. 6, the method for implementing the multicast stream forwarding includes:

Step 601: The routing bridge (RB) acquires and stores the allocation information of the multicast tree of the routing bridge group (RBv) in the downlink direction, and the multicast tree is allocated in the downlink direction in each link aggregation group (LAG) or A distribution routing bridge (RB) in a local area network (LAN), and a multicast tree in the downlink direction includes an active forwarding device and a plurality of alternate forwarding devices;

In this embodiment, the allocation information of the multicast tree acquired and stored by each routing bridge (RB) in the RBv includes allocation information of the multicast tree of itself and other RBs.

Specifically, a multicast tree explicitly or implicitly allocates a plurality of alternate forwarding devices in the downlink direction. Step 602: Each routing bridge (RB) forwards the downlink multicast stream according to the allocation information of the multicast tree in the downlink direction.

In the above embodiment, the multicast tree is allocated in the uplink direction and the downlink direction between the routing bridges (RBs) of the routing bridge group (RBv), so that the downlink routing bridge or its corresponding chain can be made. When the path fails, the forwarding of the upstream multicast stream is not affected. In addition, the multicast tree is allocated in each link aggregation group (LAG) or local area network (LAN) in the downlink direction, which may make one of them. When the link of the LAG or the LAN fails, the multicast stream forwarding of other LAGs or LANs is not affected. The routing bridge (RB) allocated by the multicast tree in the downlink direction includes one active forwarding device and several standby forwarding devices. When the primary forwarding device or its corresponding link fails, the backup forwarding device of one of the affected multicast trees is upgraded to the primary forwarding device of the multicast tree to implement multicast forwarding, thereby implementing Fast protection of the routing bridge group (RBv).

In this embodiment, when the routing bridge (RB) or its corresponding link fails, the fault-related node sends fault information to an alternate forwarding device of the affected multicast tree in the RBv; and receives the fault information. The forwarding device is upgraded to the primary forwarding device of the affected multicast tree.

A plurality of alternate forwarding devices are allocated in a multicast tree. Therefore, the above fault information is unicast to notify one of the alternate forwarding devices to be upgraded to the primary forwarding device of the multicast tree. The fault-related node determines to upgrade the standby forwarding device of the primary forwarding device based on the load balancing policy according to the allocation information of the multicast tree.

In the foregoing embodiments, after the current primary forwarding device receives the fault recovery information of the original primary forwarding device or its corresponding link, the active forwarding device reverts to the standby forwarding device.

One of ordinary skill in the art will appreciate that all or a portion of the above steps may be accomplished by a program instructing the associated hardware, such as a read-only memory, a magnetic disk, or an optical disk. Alternatively, all or part of the steps of the above embodiments may also be implemented using one or more integrated circuits. Correspondingly, each module/unit in the above embodiment may be implemented in the form of hardware or in the form of a software function module. The invention is not limited to any specific form of combination of hardware and software.

The embodiments of the invention are further described below in conjunction with the accompanying drawings and application examples.

Application example one

Taking Figure 7 as an example, two LAGs are deployed under one RBv, and the entire TRILL network calculates 5 points and ¹ is Tl, T2, T3, T4 and T5.

For the uplink direction, the multicast tree is allocated between the RBs of the RBv according to the mechanism of the existing CMT document. It is assumed that the result of the RB1 allocation is {Tl, T4}, and the result of the RB2 allocation is {T2, T5}, and the result of the RB3 allocation is { T3} ; For the downlink direction, the multicast tree is allocated within the LAG of the RBv. If all RBs of the RBv are also RBs of a LAG of the RBv, the same allocation result as the uplink allocation may be used for the LAG, and the primary forwarding device of the RB in the downlink direction of the LAG in each LAG may also be used. Different multicast tree forwarding roles are allocated to the LAGs in the downlink direction. When the RBv receives the downlink multicast stream from a multicast tree, the allocation result is the same for the multicast tree in the uplink direction and the multicast tree in the downlink direction. The processing on the RB follows the existing CMT document; and for the case where the allocation result is different, whether the RB needs to forward the received downlink traffic to the downlink needs to be changed, that is, according to the multicast tree allocated in each LAG. The forwarding role in the downstream direction determines whether to forward. For details, see the last two application examples in this document.

Each RB advertises the allocated multicast tree information to other RBs to implement information synchronization between the RBs in the RBv. This can be synchronized via ESADK End Station Address Distribution Information, protocol or tunnel technology, but not limited to ESADI or Channel.

Application example 2

In the application example 2, the RB or RB link failure notification is performed in the multicast mode. Because the failure is advertised through multicast, multiple RBs will receive the announcement. In order to prevent multiple RBs from being a forwarding device of a multicast tree at the same time, only one RB can be allocated as an alternate forwarder of a multicast tree when multicast tree allocation is performed in the downlink direction.

Specifically, the primary and the alternate forwarding devices of the multicast trees are explicitly specified on each RB, and the allocation information of the uplink and downlink multicast trees as shown in FIG. 3 is in the LAG1. RB1 is the primary forwarding device of T1 and T4, and is the alternate forwarding device of T3. In LAG2, RB1 is the primary forwarding device of T2 and T3, and is the standby forwarding device of T1.

In this mode, the primary and backup forwarding devices of a multicast tree need to be explicitly specified. For each multicast tree, only one RB in the LAG of the RB v is its primary forwarding device and standby. Forward the device. After receiving the multicast advertisement message, the RB can determine whether the received multicast stream needs to be forwarded according to the locally saved information and the fault information carried in the advertisement message.

When an RB receives the multicast stream in the downlink direction of the multicast tree, it determines whether the RB is the primary forwarding device in the downlink direction of the LAG, and if so, from the corresponding LAG interface. Export the traffic; otherwise, discard the traffic;

As shown in Figure 7, when the flow forwarded by RB4 along the T1 multicast tree reaches RBv, RBI, RB2, and RB3 can receive the flow. However, since RB1 is the primary forwarding device of T1 under LAG1, RB1 strips the flow. After the TRILL header is removed, it is forwarded to CE1 through the interface in LAG1, and RB2 and RB3 do not forward the flow through the interface in LAG1. At the same time, RB2 is the primary forwarding device of T1 under LAG2, then RB2 will also strip the stream. The TRILL header is forwarded to CE2 through the interface in LAG2, and RB1 and RB3 do not forward the flow through the interface in LAG2.

For each RB, when the RB fails or the corresponding RB link under a LAG fails, the node related to the fault (specifically, if it is an RB failure, the announcement message is sent by the CE; if it is the RB link failure) If the RB and the CE corresponding to the link are aware of the fault, then the RB or the CE or both can advertise. The fault information is in the RBv through the multicast message (which can be a control plane message or a data plane OAM message, etc.). Flooding, the message indicates which RB is faulty on the interface under which LAG. In other words, the stream on the multicast tree in which the RB is the primary forwarding device cannot be forwarded to the CE through the interface of the RB in the LAG. After receiving the advertisement message, the other RBs determine, according to the locally stored multicast tree allocation information and the received advertisement message, whether it is the backup forwarding device of the multicast tree specified in the message under the LAG. The RB is upgraded to the primary forwarding device of the multicast tree in the LAG, and the multicast stream is forwarded through the corresponding LAG interface; otherwise, the received multicast stream is discarded.

As shown in Figure 7, when the RB link of LAG1 fails between RB1-CE1, the multicast stream affected by the fault is the forwarding of T1 and T4 on the LAG1 interface. Then, after RB2 and RB3 receive the multicast message and learn the fault, RB2 judges that it is the alternate forwarding device of T1 and T4 of RB1 under LAG1, then RB2 upgrades itself to the main T1 and T4 under LAG1. Using the forwarding device, and forwarding the traffic on T1 and T4 to CE1 through the interface under LAG1, RB3 determines that it is not the alternate forwarding device of the primary multicast tree T1 and T4 allocated by RB1 under LAG1, and will not receive it. The incoming stream is forwarded through the interface of LAG1. The downlink forwarding of each RB on the interface to which LAG2 belongs is not affected.

Application example 3

Specify which multicast tree's primary forwarding device is only on each RB In the scenario of RB or RB link failure notification in unicast mode. In this case, only one RB will receive the fault advertised by the faulty RB, and multiple RBs will not appear as the primary forwarding device of a multicast tree. The application instance can display the settings or implicitly default the rest. Or all RBs are alternate forwarding devices of a multicast tree, so that no matter which RB fails, the fast protection in the RBv can be quickly realized.

As shown in Figure 8, the allocation information of the allocated uplink and downlink multicast trees, in LAG1, assigns the binding relationship of {LAGl, Tl, T4} to RB1, then it can be known that RB1 is T1 and T4 is in LAG1. The internal forwarding device can be used as the backup forwarding device of T2, Τ3, and Τ5; in addition, {LAG1, Τ2, Τ5} is also allocated for RB2, then it can be known that RB2 is the primary forwarding device of Τ2 and Τ5 in LAG1. It can be used as an alternate forwarding device for T1, Τ3, and Τ4. By assigning {LAG1, Τ3} to RB3, it can be known that RB3 is the primary forwarding device of Τ3 in LAG1, and can be used as an alternate forwarding device for T1, Τ2, Τ4, and Τ5, which can be seen that Τ3 is under LAG1. There are two standby forwarding devices RB1 and RB2. When the primary forwarding device of Τ3 fails, RB1 or RB2 needs to be notified by unicast failure notification mode, and one of them is forwarded by protection. That is, in this mode, when an RB link fails, the fault-related node (such as the RB of the faulty link) automatically selects one of the devices from the remaining RBs of the RBv or randomly based on the load balancing principle. The forwarding device of the multicast tree affected by the fault and unicasts the fault information. The device that receives the unicast failure information automatically upgrades itself to the primary forwarding device of the multicast tree affected by the failure. It is not necessary to use Application Instance 2 to determine whether forwarding is required based on locally saved information and received notification messages. When an RB receives the multicast stream in the downlink direction of the multicast tree, it is determined whether the RB is the primary forwarding device in the downlink direction of the LAG, and if yes, the traffic is derived from the corresponding LAG interface. Otherwise, discard the traffic;

As shown in Figure 8, when the RB4 flows along the T3 multicast tree to the RBv, RBI, RB2, and RB3 can receive the flow. However, since RB3 is the primary forwarding device of T3 under LAG1, RB3 will strip the flow. After the TRILL header is removed, it is forwarded to CE1 through the interface of LAG1, and RB1 and RB2 do not forward the flow through the interface of LAG1. Meanwhile, RB1 is the primary forwarding device of T3 under LAG2, and RB1 also strips the flow from the TRILL header. Afterwards, it is forwarded to CE2 through the interface of LAG2, and RB2 and RB3 will not forward the flow through the interface of LAG2.

For each RB, when the RB or RB link of a LAG of the RB fails, The barrier-related node sends an advertisement message to one of the remaining RBs in a unicast manner. When the RB receives the message, it knows that it needs to be the primary forwarding device of the multicast tree affected by the fault.

As shown in Figure 8, when the RB link of LAG1 fails between RB3-CE1, RB3 chooses to unicast the fault message to RB2, and RB2 upgrades itself to the primary forwarding device of multicast tree T3 allocated under LAG1. When RB2 receives the multicast stream from T3, it will derive the multicast stream from the interface of LAG1. The downlink forwarding of each RB on the interface to which LAG2 belongs is not affected.

Application example four

This application example focuses on the process after failure recovery:

As shown in Figure 7, after the link of RB1-CE1 on LAG1 is restored, RB1 or CE1 advertises the fault recovery information to other RBs. This can be ESADI to advertise or OAM messages.

RB2 receives the above-mentioned notification message, and combines the locally saved information to determine that RB1 is the primary forwarding device of T1 and T4 in LAG1, then RB2 is rolled back to the alternate forwarding device of T1 and T4 in LAG1, and the information is Announced to RB1.

The RBI receives the RB2 reply and assigns RB1 as the primary forwarding device for T1 and T4 in LAG1.

Corresponding to the foregoing method, the present invention further provides a routing bridge (RB). As shown in FIG. 9, the routing bridge includes:

An allocation information storage unit, configured to store allocation information of a multicast tree, where the multicast tree is separately allocated in an uplink direction and a downlink direction;

The multicast stream forwarding unit is configured to forward the uplink and/or downlink multicast stream to the multicast tree. Optionally, the multicast stream forwarding unit includes:

a multicast stream receiving module, configured to receive a multicast stream in an uplink or a downlink direction of the multicast tree; and a multicast stream determining module, configured to determine, according to the allocation information, whether the forwarding device is a forwarding device of the multicast tree in a corresponding direction; The multicast stream forwarding module is configured to forward the multicast stream when it is determined to be a forwarding device of the multicast tree, or discard the multicast stream.

Preferably, the multicast tree is allocated in a downlink direction in each link aggregation group (LAG) or a local area network (LAN).

Preferably, the RB allocated by the multicast tree in the downlink direction includes one primary forwarding device and one or several standby forwarding devices, and the multicast flow determining module determines the primary forwarding device of the multicast tree in the downlink direction. When the multicast stream forwarding module determines that it is the primary forwarding device of the multicast tree, the multicast stream forwarding module forwards the multicast stream. Otherwise, the multicast stream is discarded.

As shown in FIG. 10, the routing bridge (RB) further includes:

An information receiving unit, configured to receive fault information when the routing bridge (RB) or its corresponding link fails;

The status update unit updates the active forwarding device of the multicast tree to the backup forwarding device that receives the fault information and the routing bridge (RB) is the affected multicast tree.

Specifically, the RB allocated in the downlink direction of the multicast tree includes an active forwarding device and a standby forwarding device, and the fault information is sent by the fault-related node in a multicast manner.

The allocation information stored by the allocation information storage unit includes only the allocation information of the multicast tree of the own, and one multicast tree explicitly allocates an alternate forwarding device in the downlink direction, and the fault information is multicast by the node related to the fault. Sending, the multicast fault information includes the affected multicast tree information or the LAG or LAN information of the affected multicast tree; or

The allocation information of the multicast tree stored by the allocation information storage unit includes allocation information of the multicast tree of itself and other RBs, and one multicast tree explicitly allocates one or several standby forwarding devices in the downlink direction, and the fault information The fault-related node sends the unicast to the alternate forwarding device.

Optionally, a multicast tree explicitly or implicitly allocates a plurality of standby forwarding devices in the downlink direction, and the allocation information of the multicast tree stored by the allocation information storage unit includes allocation information of the multicast tree of itself and other RBs. The fault information is sent in a unicast manner, and the RB that receives the fault information is one of the alternate forwarding devices of the affected multicast tree.

Preferably, as shown in FIG. 11, the routing bridge further includes a fault information sending unit, configured to The unicast object sent by the fault information is determined according to the allocation information of the multicast tree randomly or based on a load balancing policy.

Specifically, the allocation information storage unit uses the same multicast tree allocation algorithm or configuration to obtain the allocation information of the multicast tree of itself and other RBs, or configures the allocation information of the multicast tree of its own, and acquires other RBs interactively. Multicast tree information.

Understandably, as described above, when the RB fails, the failure information is sent by the CE. Preferably, the information receiving unit is further configured to: when the routing bridge is the current primary forwarding device, receive failure recovery information of the original primary forwarding device or its corresponding link;

The status update unit updates to the alternate forwarding device of the multicast tree after receiving the failure recovery information.

In the method and the routing bridge of the present invention, the multicast tree in the allocation information of the multicast tree is allocated in the uplink direction and the downlink direction respectively. Therefore, when the downlink direction fails, the multicast forwarding in the uplink direction is not affected. In the RBv scenario, when the access link or the access RB fails, the remote RB does not need to detect the fault, and the RB in the RBv quickly protects the multicast data in the downlink direction to ensure the protection. The end system normally receives traffic. In addition, the invention provides a scenario in which only one LAG is deployed in one RBv, and a scenario in which multiple LAGs are deployed in one RBv, that is, the forwarding allocation of the multicast tree in the downlink direction on the RBv is at the granularity of the LAG. When an RB is faulty on an RB interface belonging to a LAG, only the multicast stream affected by the fault is forwarded by the standby forwarding device, thereby achieving high reliability access of the end system to the TRILL network.

The unit/module described in the embodiment of the present invention is only an example of dividing according to its function. It can be understood that one or more other functional divisions can be given by those skilled in the art in the case of implementing the same function. In the specific application, any one or more of the functional modules may be implemented by using one functional entity device or unit, and the above transformation manner is undeniably within the protection scope of the present application.

The embodiments described herein are only a part of the embodiments of the invention, and not all of the embodiments. Based on the embodiments of the present invention, those skilled in the art are not required to make creative labor. All other embodiments obtained below are within the scope of the invention.

Industrial applicability

Claims

claims

1. A method for implementing multicast flow forwarding, including: Each routing bridge (RB) forwards the received uplink or downlink multicast flow according to the allocation information of the multicast tree. The multicast tree is based on the uplink direction and the downlink direction. assigned separately.

2. The method according to claim 1, wherein the process of forwarding the received uplink or downlink multicast flow by each routing bridge (RB) according to the allocation information of the multicast tree includes:

Receive multicast streams in the upstream or downstream direction of the multicast tree;

Determine whether it is the forwarding device of the multicast tree in the corresponding direction according to the allocation information; if so, forward the multicast flow, otherwise discard the multicast flow.

3. The method of claim 1, wherein the multicast tree is allocated in each link aggregation group (LAG) or local area network (LAN) in the downlink direction.

4. The method of claim 1, wherein a routing bridge (RB) allocated by a multicast tree in the downlink direction includes a primary forwarding device and one or several backup forwarding devices, and the routing bridge (RB) ) forwards the multicast stream of the multicast tree when it is determined to be the active forwarding device of the multicast tree; otherwise, the multicast stream is discarded.

5. The method of claim 4, wherein the allocation information of the multicast tree only includes the allocation information of the multicast tree of the current RB, and a multicast tree in the downlink direction explicitly allocates a backup forwarding device, and the routing network When a bridge (RB) or its corresponding link fails, the fault-related node multicasts fault information to all other routing bridges (RB) in the RBv; the multicast fault information includes the affected multicast tree information. Or/and the LAG or LAN information where the affected multicast tree is located; the backup forwarding device that receives the fault information and is the affected multicast tree is upgraded to the active forwarding device of the multicast tree.

6. The method of claim 4, wherein the allocation information of the multicast tree includes the allocation information of the multicast tree of the current RB and other RBs, and a multicast tree in the downlink direction explicitly allocates a backup forwarding device. When , the routing bridge (RB) or its corresponding link fails, and the node related to the failure sends fault information to the backup forwarding device of the affected multicast tree; the node that receives the fault information The backup forwarding device is upgraded to the active forwarding device of the multicast tree.

7. The method according to claim 4, wherein the allocation information of the multicast tree includes the allocation information of the multicast tree of the current RB and other RBs. In the downlink direction, one multicast tree explicitly or implicitly allocates several RBs. When there is a backup forwarding device, the routing bridge (RB) or its corresponding link fails, and the fault-related node sends fault information to a backup forwarding device of the affected multicast tree in the RBv; the backup device that receives the fault information The forwarding device is upgraded to the active forwarding device of the affected multicast tree.

8. The method of claim 7, wherein the node related to the fault determines the backup forwarding device to be upgraded to the main forwarding device randomly or based on the load balancing policy according to the allocation information of the multicast tree.

9. The method according to claim 6 or 7, wherein each routing bridge (RB) uses the same multicast tree allocation algorithm or configuration to obtain the allocation information of the multicast tree, or configures to obtain the allocation information of its own multicast tree. After allocating information, the multicast tree information of other RBs is obtained through interaction.

10. The method of claim 5, wherein the current active forwarding device switches to the backup forwarding device again after receiving the fault recovery information of the original active forwarding device or its corresponding link.

11. A routing bridge (RB), which includes:

The allocation information storage unit is configured to: store the allocation information of the multicast tree, where the multicast tree is allocated in the uplink direction and the downlink direction respectively;

The multicast flow forwarding unit is configured to: forward the uplink and/or downlink multicast flow according to the allocation information of the multicast tree.

12. The routing bridge (RB) as claimed in claim 11, wherein the multicast stream forwarding unit includes:

The multicast stream receiving module is configured to: receive the multicast stream in the upstream or downstream direction on the multicast tree; the multicast stream judgment module is configured to: determine whether it is the multicast stream in the corresponding direction of the multicast tree based on the allocation information. forwarding equipment;

Multicast stream forwarding module, set to: When it is determined to be a forwarding device on the multicast tree, forward the group stream, otherwise the multicast stream is discarded.

13. The routing bridge (RB) as claimed in claim 11, wherein the multicast tree is allocated in each link aggregation group (LAG) or local area network (LAN) in the downstream direction.

14. The routing bridge (RB) according to claim 11, wherein the multicast flow judgment module is configured to: the RB allocated by a multicast tree in the downlink direction includes a primary forwarding device and one or several backup devices. The forwarding device determines whether it is the primary forwarding device of the multicast tree in the downlink direction. When the multicast stream forwarding module determines that it is the primary forwarding device of the multicast tree, it forwards the multicast stream, otherwise it discards the group. Streaming.

15. The routing bridge (RB) as claimed in claim 14, wherein the routing bridge (RB) further includes:

The information receiving unit is set to: receive fault information when the routing bridge (RB) or its corresponding link fails;

The status update unit is set to: after receiving the fault information and the routing bridge (RB) is the backup forwarding device of the affected multicast tree, update the main forwarding device of the multicast tree.

16. The routing bridge (RB) as claimed in claim 15, wherein the allocation information stored in the allocation information storage unit only includes allocation information of its own multicast tree, and each multicast tree explicitly allocates a multicast tree in the downlink direction. Backup forwarding device, the fault information is sent by the node related to the fault through multicast, and the multicast fault information includes the affected multicast tree information or the LAG or LAN information where the affected multicast tree is located; or , the allocation information of the multicast tree stored in the allocation information storage unit includes the allocation information of the multicast tree of itself and other RBs. In the downlink direction, a multicast tree is explicitly allocated a backup forwarding device, and the fault information is determined by the fault. The relevant nodes are sent to the backup forwarding device through unicast.

17. The routing bridge (RB) according to claim 15, wherein the allocation information of the multicast tree stored in the allocation information storage unit includes the allocation information of the multicast tree of itself and other RBs, one in the downlink direction. The multicast tree explicitly or implicitly allocates several backup forwarding devices, the fault information is sent in a unicast manner, and the RB that receives the fault information is one of the backup forwarding devices in the affected multicast tree.

18. The routing bridge (RB) as claimed in claim 11, wherein the routing bridge (RB) further includes:

The allocation information synchronization unit is set to: synchronize the allocation information of their respective multicast trees with RBs of the same RBv;

The fault information is sent by the fault-related nodes in a unicast manner according to the synchronized allocation information.

19. The routing bridge (RB) according to claim 17, wherein the routing bridge further includes a fault information sending unit, configured to: determine the fault randomly according to the allocation information of the multicast tree or based on a load balancing policy The unicast object to which the information is sent.

20. The routing bridge according to claim 16 or 17, wherein the allocation information storage unit is configured to: obtain the allocation information of the multicast tree of itself and other RBs using the same multicast tree allocation algorithm or configuration, or Configure to obtain the distribution information of its own multicast tree, and interactively obtain the multicast tree information of other RBs.

21. The routing bridge (RB) as claimed in claim 15, wherein the information receiving unit is further configured to: when the routing bridge is the current active forwarding device, receive the original active forwarding device or its corresponding link. Road fault recovery information;

The status update unit is set to: after receiving the fault recovery information, update the backup forwarding device of the multicast tree.