WO2006017982A1

WO2006017982A1 - A rerouting method in the multi-protocol label switch network

Info

Publication number: WO2006017982A1
Application number: PCT/CN2005/001264
Authority: WO
Inventors: Bin Li; Guofeng Xue; Yang Cao
Original assignee: Huawei Technologies Co., Ltd.
Priority date: 2004-08-17
Filing date: 2005-08-15
Publication date: 2006-02-23
Also published as: CN1738288A; CN100518135C

Abstract

A rerouting method in the multi-protocol label switch MPLS network comprises, A, operate the label distribution protocol within the network, and each router in the network records the FEC-label mapping information received from all the ports, and generates the respective label forwarding information based on each FEC-label mapping information; B, the router configures the protected port, and specifies the backup port for the route corresponding to each FEC that considers this port as the next hop; for the route corresponding to one of the FECs, the protected port is the main port; C, the router generates the local label forwarding table based on the main label forwarding information and the backup :Label forwarding information; D, during the data transmission, if the router detects the fault of the main port, the data is routed directly to the backup port to continue the data transmission according to the backup forwarding information in the label local label forwarding table. The usage of the invention can save the time for rerouting, and has small device overhead, and this invention also can be achieved locally, and has well compatibility.

Description

Method for rerouting in multi-protocol label switching network

The present invention relates to routing techniques in a Multi-Protocol Label Switching (MPLS) network, and more particularly to a method of re-routing in a Multi-Protocol Label Switching (MPLS) network. Background of the invention

Muxtiprotocol Label Switch (MPLS) was originally proposed to improve the forwarding speed of routers. Because MPLS is very important in traffic IP (Traffic Engeering) and virtual network (VPN) in current IP networks. Performance in two technologies. MPLS has increasingly become an important standard for expanding the scale of IP networks.

The key to the MPLS protocol is the introduction of the concept of a label. MPLS classifies packets, uses tags to identify packets that belong to the same forwarding type, and uses labels in the MPLS network as the sole identifier for determining forwarding operations.

Referring to Figure 1, Figure 1 is a schematic diagram of the network structure of MPLS. MPLS can be logically divided into a Label Switching Edge Router (LER) and a Label Switching Router (LSR). The ingress LER provides a mapping of the Forwarding Equivalence Class (FEC) and the label, and the egress LER provides the label removal function, and the LSR is the core switch of the MPLS network, which provides Label Swapping and label distribution. Features.

The MPLS connection is called a Label Switching Path (LSP). Each node forwards data packets according to Next Hop Label Forwarding Entry (NHLFE). The NHLFE contains the following contents: (1) Next Hop of the data packet (Next Hop)

(2) DDL package used when forwarding data packets

(3) How to encode the label stack used when forwarding data packets

(4) For the operation of the data packet label stack, possible operations are:

1 Replace the label at the top of the grouped label stack with a new one

2 pop-up label top

3 Replace the label at the top of the group label stack with a new label and push in one or more new labels

When data packets enter MPLS, they are first classified into different FECs, and then these FECs are mapped to NHLFE, thereby achieving the purpose of forwarding data packets by label switching. There are two mappings between FEC and NHLFE, Incoming Label Map (ILM) and FEC to NHFLE mapping (FTN). The former is used to forward the already labeled packet (Core LSR), and the latter is used to forward the packet. Labeled packets (Edge LSR, LER), but after forwarding, the packets are already tagged.

To establish an LSP, you must also use the label distribution protocol, mainly LDP ( Label Distribution Protocol ) and RSVP-TE ( Resource reSerVe Protocol Traffic Engineer extension ).

The LDP protocol includes a set of messages and procedures for establishing an LSP between LSRs. The LDP protocol directly maps the routing information of the network layer to the LSP established at the data link layer. An LSP may be established between two adjacent LSRs, or may be through an entire routing area (including multiple LSRs).

The label is assigned to specify a label to be bound to an FEC, and distribution refers to the process of notifying upstream or downstream of this binding. Label distribution and distribution are done under the control of the LDP protocol.

There are two ways to assign and distribute labels: 1. The LSR allocates FEC-label bindings to other LSRs that explicitly propose label requests. This method is called Downstream On Demand label distribution.

2. The MPLS architecture also allows the LSR to actively allocate FEC-label bindings to other LSRs that do not request a label. This method is called Downstream Unsolicited label distribution.

The main difference between the downstream autonomous label advertisement and the downstream on-demand label advertisement method is which LSR is responsible for initiating the label mapping request and the label mapping notification process.

Tag control also has an orderly and independent 'control mode' two ways:

1. When the LSP control is used, the LSR can send the label mapping message to the upstream only when the LSR receives the specific FEC-label mapping message of the next hop of the specific FEC or the LSR is the egress node of the LSP. If the LSR is neither an egress node for a particular FEC nor a label binding for a particular FEC, then the LSR must wait downstream before receiving a FEC-tag binding for a particular FEC and returning a specific FEC-label binding message upstream. A specific FEC-tag response message for the LSR.

2. When using independent LSP control, each LSR can advertise the label mapping to its associated LSR at any time. For example, when working in an independent downstream on-demand label distribution control mode, the LSR can immediately respond to the upstream label request message without waiting for the label mapping message from the next hop LSR. When working in an independent downstream autonomous label distribution control mode, the LSR can advertise a specific FEC-tag mapping message to the LSR connected to it as long as the LSR is ready for label forwarding for a particular FEC. The use of independent label control allows the LSR to forward the label mapping message upstream before receiving the downstream label mapping message.

Whether the LSR uses independent or ordered control determines the behavior of the LSR during the LSP establishment process. As a configurable option, the LSR can support both control methods. Label retention also has two modes: conservative label retention and free label retention: In the downstream autonomous label distribution mode, label mapping messages for all routes can be received from any neighboring LSR. When the conservative label hold mode is used, only the FEC-tag binding for data forwarding is retained; that is, the received FEC-tag binding comes from the next hop LSR of the route.

In the downstream on-demand label distribution mode, the LSR only sends a label request message to the FEC next hop LSR. Since downstream on-demand label distribution is mainly used in environments with limited label resources, such as ATM switches with limited cross-connect space, downstream on-demand label distribution usually uses conservative label retention. The advantage of conservative label retention is that only tags for data forwarding are assigned and maintained. This is very important for LSRs with limited label resources, such as ATM switches. One disadvantage of the conservative label hold mode is that if the route changes the next hop LSR for a particular FEC, the LSR must wait for a specific FEC-tag mapping message from the new next hop before the packet can continue to forward the label.

In the downstream autonomous label distribution mode, the LSR can receive label mapping messages for all routes from any neighboring LSR. When using the tag-holding mode, the LSR retains all tag mappings regardless of whether the sending LSR is the next hop of the particular FEC-label mapping it advertises. When using the downstream on-demand label distribution method, the LSR may choose to send a label request message to all neighboring LSRs for all known address prefixes. Downstream on-demand label distribution is typically used for devices such as ATM switches.

The main advantage of the free label hold mode is that the LSR can respond quickly to route changes, mainly because tag mapping already exists. The main disadvantage of the free label retention method is that the currently unneeded label mapping also needs to be assigned and maintained.

In an MPLS network, when a node or link on the LSP fails, MPLS fast reroute technology can be used to switch the LSP to the previously established local backup LSP. On, protecting the LSP is not affected by link/node failure.

The protection of the network service by the MPLS fast re-routing technology is shown in Figure 2. Figure 2 is a schematic diagram of the LSP path of the MPLS fast re-routing in the prior art;

The primary LSP path is A-B-C-D-E. The path AGC is used to protect the node B and related links, the path BGD is used to protect the node C and related links, the path CFE is used to protect the node D and related links, and the path DFE is used to link the link DE Protect. For example, when there is a fault at point C, point B will switch the network traffic to B-G-D, thus reducing data loss.

Currently, there are two ways to quickly reroute: Bypass and Detour. Currently, the main use of fast rerouting in Bypass mode. In Bypass mode, a pre-configured LSP is used to protect multiple LSPs. When the link fails, the primary tunnel LSP is routed to the pre-configured LSP, and the pre-configured LSP reaches the next hop router for protection purposes.

At present, this method of MPLS fast rerouting has the following disadvantages:

1. Depending on the complex MPLS TE technology, the equipment costs are large;

2. The backup LSP needs to be explicitly specified manually. The configuration workload is large and cannot be deployed on a large scale.

3. Depending on the protocols such as OSPF-TE/ISIS-TE/RSVP-TE/CR-LDP, multiple devices need to work together, and the compatibility is poor.

4. To perform link, node, and path protection, you need to establish backup LSPs separately, which brings unnecessary overhead.

5. The backup LSP also has a fault. There is no protection mechanism. When it fails, fast reroute cannot be performed.

6. The backup LSP cannot pass through the protected link or node. The requirements are too strict. Sometimes, even if the destination is reachable, the backup LSP cannot be established. Summary of the invention

In view of this, the main object of the present invention is to provide a multi-protocol label switching.

(MPLS) A method of rerouting in a network that not only enables fast rerouting after a node failure, but also saves equipment overhead.

In order to achieve the above object, the technical solution of the present invention is specifically implemented as follows:

A multi-protocol label switching method for rerouting in an MPLS network, the method comprising the following steps:

A. Running a label distribution protocol in the network, each router in the network records FEC-tag mapping information received from all ports, and generates each label forwarding information according to each FEC-tag mapping information;

B. Configure the protected port. For the route corresponding to each FEC of the next hop, determine the backup port. For a route corresponding to FEC, the protected port is the primary port.

The label forwarding information generated by the FEC-tag mapping information received by the primary port is used as the primary label forwarding information, and the label forwarding information generated by the FEC-tag mapping information received by the standby port is the standby label forwarding information.

C. The router generates a local label forwarding table according to the primary label forwarding information and the alternate label forwarding information.

D> During data transmission, if the router detects that the primary port is faulty, it forwards the information to the alternate port according to the alternate label forwarding information in the local label forwarding table.

The label distribution protocol running in the network works as follows: downstream autonomous label distribution, ordered label control, and free label retention.

The method for determining the alternate port in step B may be: manually selecting; or excluding the active port according to the principle of the shortest path, and running the link state routing protocol to automatically calculate the backup port. The manual selection may be as follows: Specify an alternate port for the primary port, and all the FECs with the primary port as the next hop, and the alternate ports are designated ports.

The automatic calculation can be configured as follows: Configure a port to be automatically protected or configure all ports of a device to be automatically protected.

The process of automatically calculating the backup port may be performed after the normal route calculation is completed. In the case of automatically calculating the backup port, if the router detects that the backup port is faulty, you can recalculate the backup port and use the obtained alternate label forwarding information corresponding to the new backup port to replace the spare label corresponding to the original backup port in the local label forwarding table. Forward the message.

The label forwarding information is: a next hop label forwarding item generated by the router according to the FEC-tag mapping information, NHLFE;

The primary label forwarding information is: an NHLFE generated by the router according to the FEC-tag mapping information received by the primary port;

The standby label forwarding information is: the standby NHLFE generated by the router according to the FEC-tag mapping information received by the backup port.

For the ingress edge router LER, the local label forwarding table includes: FTN and NHLFE;

If the port is configured to be protected, the FTN generation method in step C is: Cl, mapping the FEC with the primary NHLFE and the standby NHLFE to generate an FTN. If the FEC is configured to protect the port with the primary NHLFE and the port is protected, the step C includes:

C2, LER traverses the FTN, finds the NHLFE next hop as the protected port entry, adds a backup NHLFE to the FTN, and backs up the next hop corresponding to the NHLFE as the determined backup port.

For the label switching router LSR, the local label forwarding table includes: an ILM and an NHLFE; If the port is configured to be protected, the method for generating the ILM in the step C is: Cl, mapping the ingress label with the primary HLFE and the standby NHLFE to generate a LIM. If the port is protected after the ILM is generated, the step C includes: C2, the LSR traverses the ILM, finds the entry whose NHLFE next hop is the protected port, adds a backup NHLFE to the ILM, and backs up the NHLFE. The next hop is the determined backup port.

The method for routing data to the alternate port to continue to transmit data may be: replacing the original label in the received data with the label corresponding to the standby port, and sending the data to the standby port according to the label.

It can be seen from the above technical solution that the method for rerouting in the MPLS network of the present invention protects a specific port by selecting a backup port in an MPLS network environment running a label distribution protocol, which is applicable to The port is all LSPs of the outbound interface. When this port fails, the device can quickly switch to the backup port, and MPLS forwarding is performed through the backup port to reach the destination.

Therefore, the method of the present invention does not need to adopt complex MPLS TE technology, and realizes distributed processing of multiple nodes, and the backup port can simultaneously implement link protection, node protection, and path protection, and does not need to separately establish backups for links, nodes, and paths. LSP, which saves equipment overhead. Moreover, since the present invention is implemented locally, it does not require the support of adjacent devices, and has good compatibility. BRIEF DESCRIPTION OF THE DRAWINGS

Figure 1 is a schematic diagram of a network structure of MPLS;

2 is a schematic diagram of an LSP path of a MPLS fast reroute in the prior art;

3 is a schematic structural diagram of an MPLS network according to a preferred embodiment of the present invention;

4 is a schematic diagram of label distribution of the embodiment shown in FIG. 3; 5 is a schematic diagram of label forwarding of R1 in the embodiment shown in FIG. 3; FIG. 6 is a schematic diagram of label forwarding of R2 in the embodiment shown in FIG.

FIG. 7 is a schematic diagram of rerouting after R2 failure in the embodiment shown in FIG. 3. Mode for carrying out the invention

The present invention will be further described in detail below with reference to the accompanying drawings.

The present invention performs protection backup on a specific port by selecting a backup port in an MPLS network environment running the LDP protocol. It is applicable to all LSPs that use the port as the outbound interface. When the port is faulty, the device can quickly switch to Backup port, MPLS forwarding through the backup port to reach the destination.

The method for rerouting in the MPLS network of the present invention mainly includes the following steps:

D. During the data transmission process, if the router detects that the primary port is faulty, it forwards the information according to the alternate label in the local label forwarding table, and directly routes to the alternate port to continue the number of transmissions. According to.

A preferred embodiment will be described below.

Referring to FIG. 3, FIG. 3 is a schematic structural diagram of an MPLS network according to a preferred embodiment of the present invention. Where R1 is the ingress LER, R2, R3, and R4 are LSRs, and R5 is the egress LER.

In this embodiment, the LDP protocol is first run on the MPLS network, and its working mode is downstream autonomous (DU) label distribution, ordered label control, and free label maintenance.

In the DU label distribution mode, the egress LER does not need to make a label assignment request from the upstream LSR, and actively allocates the FEC-tag mapping, that is, the binding, and sends the message to the upstream. Therefore, in this embodiment, each router receives FEC-tag mapping information sent by multiple next hop ports.

Because in the free label hold mode, the LSR can receive the label mapping message for the FEC from any neighboring LSR, regardless of whether the LSR that sent the message is the route corresponding to the FEC in the specific FEC-label mapping it advertises. One hop, the LSR retains all label mappings. Therefore, the router in this embodiment records all received FEC-tag mapping information.

Referring to FIG. 4, FIG. 4 is a schematic diagram of label allocation of the embodiment shown in FIG. Among them, R1 arrives at R5 with two paths R1-R2-R5 and R1-R3-R4-R5, and R5 initiates a multi-label mapping message upstream. R5, R4, R3, and R2 respectively send FEC-tag mapping information to the previous hop, which contains the label allocated by itself for the previous hop.

Finally, as shown in Figure 4: R2 and R3 respectively assign R1 to R5 tags L21 and L31, R5 and R3 respectively assign R2 to R5 tags L52 and L32, R4 and R2 respectively assign R3 to R5 The labels L43 and L23, R5 assign R4 a label L54 to R5.

In this embodiment, R1 generates two NHLFEs according to the FEC-tag mapping information received from R2 and R3, and R2 also generates the FEC-tag mapping information received from R5 and R3. Two NHLFEs, R3 also generate two NHLFEs based on the FEC-tag mapping information received from R4 and R2, and R4 generates an NHLFEo based on the FEC-tag mapping information received from R5.

Then, the router configures the protected port. For each FEC corresponding route that uses this port as the next hop, the backup port is determined. For an FEC (route), the protected port is the primary port.

For example, for a certain FEC, R1-R2-R5 is the route corresponding to the FEC. For R1 and R2, it is determined as the primary port. The NHLFE generated based on the FEC-label mapping information sent by R2 is used as the primary NHLFE, and R3 is determined. For the standby port, the NHLFE generated according to the FEC-label mapping information sent by R3 is the standby NHLFEo. For R2, R5 is determined as the primary port. The NHLFE generated according to the FEC-label mapping information sent by R5 is used as the primary NHLFE, and R3 is determined as the standby. Port, the NHLFE generated according to the FEC-tag mapping information sent by R3 is the standby NHLFE.

The method of determining the alternate port is to specify one device port of the LSR as another device port, that is, the backup port of the protected port. These two ports can be either physical ports or logical ports, provided that LDP runs on this port.

The backup port can be selected manually or automatically. The manual selection method can be as follows: Specify an alternate port for the primary port, and all the FECs with the primary port as the next hop, and the backup ports are designated ports.

In complex cases, the manually selected port may not be optimal, or the configuration is heavy, so it is recommended to use the automatic calculation method to configure. The automatic calculation method calculates the backup port according to the principle of the shortest path when specifying the port to be protected.

Automatic calculation can be implemented by the device running link state routing protocol, such as OSPF, IS-IS, and the protocol does not need to be extended. According to the criteria for automatic calculation, the existing protocol can be used. The criteria for automatic calculation are: discovery from local to destination, but not through protected The shortest path of the port is to exclude the information of the protected port in the link state database, that is, to simulate the interruption of the protected port and calculate the shortest path.

There are two ways to configure the automatic calculation: one is to protect the port, and the other is to protect all the ports.

In the case of automatic calculation, as long as the destination is reachable, the backup port can always be found for protection. If the calculated backup port fails later, the new backup port can be recalculated. After the new backup port is calculated, the backup label forwarding information corresponding to the obtained new backup port is used instead of the backup label forwarding information corresponding to the original backup port in the local label forwarding table.

Since the automatic calculation consumes CPU resources, in order to prevent this calculation from slowing down the route convergence speed, the automatic calculation of the backup port is performed after the normal route calculation is completed. In this way, the calculation of the backup port is performed when the CPU is idle, and the ordinary SPF algorithm can be used to meet the requirements. It is not necessary to enhance the CPU capability or adopt a complicated algorithm, and the CPU load is low.

In this embodiment, the backup port can be manually configured or automatically generated. When the backup port is manually designated, the configuration workload is reduced compared with manually specifying the backup LSP. The automatic generation does not require manual explicit designation, which further reduces the The amount of work deployed.

In this embodiment, R1-R5 each maintains a local label forwarding table. When the port backup is not implemented, each label forwarding table has only one FTN/LIM and one NHLFE, and the HLFE is the FEC sent according to the next hop of the FEC route. - Tag mapping information generated. After the port backup is implemented, if the next hop of a label forwarding table is a protected port, an alternate NHLFE is added to the table. The standby NHLFE is generated based on the FEC-tag mapping information sent by the backup port.

The local label forwarding table maintained by the ingress LER and the intermediate LSR is slightly different, and is described in detail below.

For the ingress LER, its label forwarding table consists of two segments, FTN and NHLFE, which is a mapping of FEC to NHLFE. If the FTN is generated after the port backup configuration, the normal next hop of the FTN is determined by the IGP shortest path to the FEC, and the primary NHLFE is generated according to the FEC-tag mapping information sent by the next hop of the FEC route. When a new FTN is created, the FEC is mapped to the primary NHLFE and the backup NHLFE to generate an FTN. Thus, the FTN, the primary NHLFE, and the alternate NHLFE form a local label forwarding table.

If the FTN is generated before the port backup configuration, the FEC is mapped to the primary NHLFE. After the FTN is generated, the LER traverses the FTN table to find the entry whose NHLFE next hop is the protected port, and adds a backup NHLFE to the FTN. Backup The next hop corresponding to NHLFE is the determined backup port. Thus, the FTN, the primary NHLFE, and the standby NHLFE form a local label forwarding table.

Referring to FIG. 5, FIG. 5 is a schematic diagram of label forwarding of R1 in the embodiment shown in FIG. The figure shows the main entries in the R1 tag forwarding table: Destination R5, Primary NHLFE R1-R2, L21, Alternate NHLFE R1-R3, L31.

For the intermediate LSR, its label forwarding table consists of two segments: ILM and NHLFE. The ILM is the mapping table between the incoming label and the NHLFE. Since both inbound and outbound tags are assigned to a specific FEC, an ILM has a unique correspondence with this FEC.

If the ILM is generated after the port backup configuration, the inbound label is mapped to the primary NHLFE and the standby NHLFE to generate UM. In this way, LIM, primary NHLFE, and standby NHLFE form a local label forwarding table.

If the ILM is generated before the port backup configuration, the ingress label is also mapped to the primary NHLFE to generate the LIM. Then the LSR traverses the LIM table to find the entry whose NHLFE next hop is the protected port, and adds a backup to the LIM. NHLFE, the next hop corresponding to the backup NHLFE is the determined backup port. In this way, LIM, primary NHLFE, and standby NHLFE form a local label forwarding table.

Referring to FIG. 6, FIG. 6 is a schematic diagram of label forwarding of R2 in the embodiment shown in FIG. Shown in the figure The main entries in the R2 label forwarding table are: Incoming label L21, primary NHLFE R2-R5, L52, standby NHLFE R2-R3, U2.

It can be seen from FIG. 6 that for the route R1-R2-R5, R2 is the penultimate hop. In fact, the label L52 assigned by the egress LER R5 belongs to a special control label, that is, R5 does not assign a substantial label to the FEC. R2, but the next hop that the backup port R3 arrives is not the LER. On this port, the FEC needs to be assigned a label. Therefore, R3 is still required to be assigned a label when generating a backup NHLFE.

In order to reroute after a device failure, R1-R5 also maintains the working status of each port: Normal/Failed. When it is detected that a port is not working properly, such as a physical link failure or manual operation, the port is closed and its status is updated immediately. During the data forwarding process, the label forwarding table can be used to obtain the next-hop primary port of the data. If the status of the packet is invalid, the switch is switched to the backup port, and the corresponding label is set, that is, the label corresponding to the alternate port is replaced. Receive the original label in the data, and send the data to the alternate port according to the label.

The data arrives at the alternate port. Since the tag is assigned by itself, there must be a corresponding tag forwarding table in it, so that the data can continue to be forwarded to the destination.

Referring to FIG. 7, FIG. 7 is a schematic diagram of rerouting after R2 failure in the embodiment shown in FIG. It is assumed that the LSP of R1-R5 is protected, the shortest path is R1-R2-R5, and the backup port is Rl-R3. On R3, the shortest path to R5 is R3-R2-R5 and the backup port is R3-R4. If R2 fails, both R1 and R3 sense this change and switch to the backup port respectively, thus forming a new LSP, R1-R3-R4-R5. As long as R1-R5 maintains reachability, multiple nodes work independently. It can protect any link and node failure on the path, that is, it can realize path protection.

It can be seen that if multiple nodes are used in the network, the port protection can be implemented, and the node and path protection can be performed at the same time. It is not necessary to separately establish backup LSPs for links, nodes and paths, and no need to add new Features. The method for implementing MPLS network re-routing by port backup of the present invention can protect the LSP established by the LDP protocol and the CR-LSP established by the RSVP-TE/CR-LDP at the same time.

The destination of the CR-LSP is the egress LER device. LDP establishes a label forwarding table to reach this LER. After the backup port is manually configured or automatically calculated, the backup next hop is added to the label forwarding table corresponding to the CR-LSP. The port is the backup port. The LDP peer whose label is the backup port is allocated to the egress LER.

The method of the present invention can also protect IP traffic. At this time, the primary next hop of the forwarding table uses IP forwarding, and the alternate next hop uses MPLS forwarding. At this time, the label forwarding table has only FTN/ILM and a standby NHLFE. .

It can be seen from the foregoing embodiments that the method for rerouting the MPLS network of the present invention adopts a method of simulating link failure and calculating a backup port in advance, thereby eliminating the route calculation time after the fault occurs and the time for reestablishing the LSP. This speeds up the rerouting. Moreover, the present invention does not require the use of complex MPLS TE technology, has low equipment overhead, can be implemented locally, does not require support from adjacent devices, and has good compatibility.

Claims

Claim

A multi-protocol label switching method for rerouting in an MPLS network, characterized in that the method comprises the following steps:

B. The router is configured with a protected port. For each FEC corresponding route that uses this port as the next hop, the backup port is determined. For a route corresponding to the FEC, the protected port is the primary port.

D. During the data transmission process, if the router detects that the primary port is faulty, it forwards the information according to the alternate label in the local label forwarding table, and directly routes to the alternate port to continue transmitting data.

The method of claim 1, wherein the label allocation protocol operating in the network operates in the following manner: downstream autonomous label distribution, ordered label control, and free label retention.

3. The method according to claim 1, wherein the method for determining the alternate port in step B is: manually selecting; or excluding the primary port according to the principle of the shortest path, and automatically calculating the running link state routing protocol. Backup port.

4. The method of claim 3, wherein the manner of manual selection is: Specify an alternate port for the primary port. All the FECs with the primary port as the next hop are the designated ports.

5. The method of claim 3, wherein the automatically calculated configuration mode is: configuring a port to be automatically protected or configuring all ports of a device to be automatically protected.

The method according to claim 3 or 5, wherein the process of automatically calculating the backup port is performed after the normal route calculation is completed.

The method according to claim 3 or 5, wherein: in the case of automatically calculating the backup port, if the router detects that the backup port is faulty, recalculating the backup port and using the obtained new backup port for the backup Label forwarding information, which replaces the alternate label forwarding information corresponding to the original backup port in the local label forwarding table.

The method of claim 1, wherein the label forwarding information is: a next hop label forwarding entry NHLFE generated by the router according to the FEC-tag mapping information; and the primary label forwarding information is : The FEC-tag mapping information received by the router host port, the generated HLFE;

9. The method according to claim 8, wherein, for the ingress edge router LER, the local label forwarding table includes: FTN (FEC and NHLFE mapping) and NHLFE; if the port has been configured to be protected, the step The method for generating the FTN in C is: Cl, mapping the FEC with the primary NHLFE and the standby NHLFE to generate an FTN.

The method according to claim 8, wherein, for the LER, the local label forwarding table includes: FTN and NHLFE;

If the port is protected after the FEC is configured with the primary NHLFE to generate the FTN, the step C includes:

C2, the LER traverses the FTN table, and finds the entry whose NHLFE next hop is the protected port. Add a backup NHLFE for this FTN, and back up the next hop corresponding to the NHLFE to the determined backup port.

The method according to claim 8, wherein, for the label switching router LSR, the local label forwarding table includes: an ILM (incoming label and out label mapping) and an NHLFE;

If the port is configured to be protected, the method for generating the ILM in the step C is: Cl, mapping the inbound label with the primary NHLFE and the standby NHLFE to generate a LIM.

The method according to claim 8, wherein, for the LSR, the local label forwarding table includes: an ILM and an NHLFE;

If the port is protected after the ILM is generated, the step C includes: C2, the LSR traverses the ILM table, searches for the entry of the NHLFE next hop as the protected port, and adds a backup NHLFE to the ILM. The next hop corresponding to the NHLFE is the determined backup port.

The method of claim 1, wherein the method for routing data to the alternate port to continue to transmit data is: replacing the original label in the received data with a label corresponding to the alternate port, and sending the data according to the label Give the alternate port.