CN101984591A

CN101984591A - Service convergence method and system in transport-multiprotocol label switching network

Info

Publication number: CN101984591A
Application number: CN2010105119790A
Authority: CN
Inventors: 杨红燕; 周鹏辉
Original assignee: ZTE Corp
Current assignee: ZTE Corp
Priority date: 2010-10-13
Filing date: 2010-10-13
Publication date: 2011-03-09
Anticipated expiration: 2030-10-13
Also published as: CN101984591B

Abstract

The invention discloses a service convergence method in a transport-multiprotocol label switching (T-MPLS) network, which comprises the following steps: a provider-edge device (PE) connected with a customer-edge device regularly scans the port status of respective customer access port, and a customer signal failure (CSF) message is transmitted to the CE-dual homed PE when the port is in a port fault status; after receiving the CSF message, the CE-dual homed PE updates the value in a master PW ACTIVATED/FAILED field among its own forwarding-table items to FAILED; and after receiving the data message to be transmitted to the far-end CE, the CE-dual homed PE transmits the data message according to the updated forwarding-table items. The invention also discloses a service convergence system in a T-MPLS network. By using the method and system, the invention can realize the rapid convergence of the service in case of a link fault, thereby meeting the requirements of the carrier Ethernet for service switching time.

Description

Method and system for transmitting service convergence in multi-protocol label switching network

Technical Field

The present invention relates to a transport multi-protocol Label Switching (T-MPLS) network technology, and more particularly, to a method and system for service convergence in a T-MPLS network.

Background

To accommodate the demands of the evolution of metropolitan area networks on transport networks, the international telecommunication union telecommunication standardization sector (ITU-T) extended the concept of Multiprotocol Label Switching (MPLS) networks to a lower transport level, which was proposed at ITU-T SG15 conference in 2006, 2 months. T-MPLS is a connection-oriented packet transport technology based on MPLS, the data forwarding plane of T-MPLS is a subset of MPLS, and the data of T-MPLS is forwarded based on T-MPLS label. T-MPLS is the application of MPLS in transport networks that simplifies some of the complex functions of the MPLS data forwarding plane, namely: unnecessary processing of Internet Protocol (IP) facing interconnection between networks is omitted, and cost of operators can be greatly reduced; the T-MPLS network has complete service expansion capability, supports the transmission of signals of different technologies and conforms to the trend of network transformation; compared with the traditional packet transport network, the T-MPLS network has the connection-oriented characteristic, and the functions of strong Operation, Administration and Maintenance (OAM) and protection recovery and the like guarantee the quality requirement of the Operation level.

In a T-MPLS network, when a user side edge device (CE) is under a network model of a dual-homed provider edge device (PE), according to a different node failure, a fast switching of a service is generally solved by using T-MPLS tunnel protection or Pseudo Wire (PW) protection when the network fails. The basic realization principle is as follows: an end-to-end T-MPLS tunnel and PW connection are established between two PE devices, a standby link is established in advance for a main link needing protection, and when the PE devices detect that the main link fails, such as node failure or link failure, flow is switched to the standby link, so that rapid switching of services is realized. Here, the node failure means that the node has a failure that does not affect the backup link.

Fig. 1 is a typical network model of CE dual-homed PE, and further details the existing tunnel protection and PW protection service switching method based on T-MPLS network in conjunction with fig. 1: in fig. 1, PE _ a is a PE that is dually homed to CE _ B, and PE _ B and PE _ C are both PEs connected to CE _ B. Setting the main link of accessing CE _ B by CE _ A as follows:

CE_A-PE_A-P_A-PE_B-CE_B；

when PE _ A or a link between PE _ A and PE _ B fails, the link accessing CE _ B by CE _ A converges to be:

CE _ A-PE _ A-P _ B-PE _ C-CE _ B; here, the failure of PE _ a means that PE _ a has a failure that does not affect the backup link.

According to the standard T-MPLS tunnel protection technology, in the actual data packet processing process, the forwarding of the data packet is determined by first querying a forwarding table, and then determining which link is used for forwarding, so that a forwarding table needs to be configured on each node in advance, and the forwarding table in the prior art is configured as follows:

the configuration of the T-MPLS service is mainly divided into an example part and a T-MPLS tunnel part, wherein the example part is mainly used for accessing a data message at one side of a private network user to one side of a public network tunnel; or, the data message at one side of the public network tunnel is accessed to one side of the private network user. The configuration near the private network user side is called Client Instance Port (CIP) configuration, and includes information accessed by the user, such as interface type, interface content, and the like; the configuration near one side of the public network tunnel is called a Virtual Instance Port (VIP) configuration, and includes information of a public network tag, a private network tag, and a next hop of the tunnel. During configuration, the tunnel and the PW are usually bound to the VIP, so as to determine information of a private network tag of a forwarding entry, a public network tag of the tunnel, and a next hop.

The configuration of the T-MPLS tunnel is on nodes, each node being a segment of the tunnel, and an egress on each node having an ingress, but not all nodes having an ingress and an egress, specifically, on the head of the tunnel, i.e. the head node, having no ingress, the missing ingress being supplemented by the customer instance access; similarly, there is no egress at the end of the tunnel, i.e., the end node, and the missing egress is also supplemented by client instance access. And configuring a tunnel protection group on a configuration node of the tunnel to protect the main tunnel. And configuring a PW protection group on the PE node to protect the main PW.

According to the configuration method, through the configuration of the example part and the tunnel part, the information of two links from PE _ A to PE _ B and from PE _ A to PE _ C in FIG. 1 is written into the forwarding table entry of the PE _ A equipment, the information of the links comprises a public network label, a private network label, tunnel information, PW information and the like, in the actual data message processing process, the judgment is carried out through inquiring the forwarding table entry, and then how to forward is determined.

In the network model shown in fig. 1, when a link between PE _ a and PE _ B or a PE _ a node fails, fast switching of services can be implemented through T-MPLS tunnel protection. When PE _ B node fails, the end-to-end protection configuration of T-MPLS tunnel starts from the tunnel head node and ends at the tunnel tail node, so when PE _ B fails, the tunnel protection group is invalid, at this time, the T-MPLS tunnel protection can not realize the rapid switching of the service, therefore, the rapid switching of the service can be realized through PW protection.

When the link between PE _ B and CE _ B fails, the convergence of service can be realized only through PW protection. However, the current carrier ethernet requires that the time for recovering normal transmission of the service is less than 50ms, and since the end-to-end service convergence rate is closely related to the number of users accessing the private network and the number of hops in the bearer network, in other words, PE _ a cannot quickly know that a link between PE _ B and CE _ B has a fault in most cases, in the existing networking, the end-to-end service convergence time cannot basically meet the requirement that the carrier ethernet is less than 50 ms.

Disclosure of Invention

In view of the above, the main objective of the present invention is to provide a method and a system for service convergence in a T-MPLS network, which can achieve fast service convergence when a link fails, so as to meet the requirements of a carrier ethernet network.

In order to achieve the purpose, the technical scheme of the invention is realized as follows:

the invention provides a method for service convergence in a T-MPLS network, which comprises the following steps:

the PE connected with the CE scans the port state of a client access port of the PE at regular time, and sends a Client Signal Failure (CSF) message to the PE of the CE dual-homing when the port state is a port fault;

after the CE dual-homed PE receives the CSF message, updating a field value of whether the primary PW takes effect in a self-forwarding table entry to be invalid;

and after receiving the data message which needs to be sent to the remote CE, the CE dual-homed PE sends the data message according to the updated forwarding table entry.

In the above solution, the timing duration of the PE connected to the CE scanning the port state of its own customer access port is not greater than 10 ms.

In the above solution, before the PE connected to the CE periodically scans the port status of its own customer access port, the method further includes:

and setting T-MPLS tunnel head node information and example information on the CE dual-homed PE, setting T-MPLS tunnel tail node information and example information on the CE connected PE, forming a forwarding table item on the CE dual-homed PE, and then setting bidirectional forwarding detection on the CE dual-homed PE and the CE connected PE.

In the above scheme, the sending the data packet according to the updated forwarding table entry includes:

the CE double-homed PE inquires an instance table in a forwarding table item, when the effective field value of the primary PW is found to be invalid, the data message is sent through the standby PW, when the data message is sent, the CE double-homed PE inquires a tunnel table according to a tunnel number bearing the PW corresponding to the standby PW in the instance table, encapsulates required information such as a public network label and a private network label in the data message according to the content corresponding to the tunnel table, and sends the encapsulated data message to a far-end CE through the standby PW.

In the above scheme, the method further comprises:

and when the primary PW is recovered to be normal and the port state of the PE connected with the CE is normal, updating the value of the effective field of the primary PW in the forwarding table entry to be effective by the PE of the CE dual homing.

In the above solution, the PE connected to the CE periodically scans the port status of its customer access port, and sends a CSF packet to the PE dual-homed to the CE when the port status is a port failure, where the method includes:

the detecting module of the PE connected with the CE scans the port state of the client access port at regular time, and triggers the CSF message sending module of the PE connected with the CE when the port state is scanned as a port fault for the first time, the CSF message sending module sends CSF messages to the PE with the CE dual homing, and meanwhile, the detecting module sets an CSF message sending identifier;

and the CSF message sending module periodically and automatically sends CSF messages to the PE with the CE dual homing.

In the above scheme, the method further comprises:

after the detection module scans that the port state is normal, the CSF message sending identification is cleared, and a notice of stopping sending the CSF message is sent to the CSF message sending module; and after receiving the notice of stopping sending the CSF message, the CSF message sending module stops sending the CSF message to the PE with the CE dual homing.

The invention also provides a system for service convergence in the T-MPLS network, which comprises: a first PE and a second PE; wherein,

the first PE is used for scanning the port state of a client access port of the first PE at regular time and sending a CSF message to the second PE when the port state is a port fault;

the second PE is used for updating the value of the field whether the main PW takes effect in the self forwarding table entry to be invalid or not after receiving the CSF message sent by the first PE; and after receiving the data message which needs to be sent to the remote CE, sending the data message according to the updated forwarding table entry.

In the foregoing solution, the system further includes a setting unit, configured to set, on the second PE, head node information and instance information of a T-MPLS tunnel, set, on the first PE, tail node information and instance information of the T-MPLS tunnel, form a forwarding entry on the second PE, and then set bidirectional forwarding detection on the first PE and the second PE.

In the foregoing solution, the second PE is further configured to update the value of the valid field of the active PW to be valid when the active PW recovers to normal and the port state of the PE connected to the CE is port normal.

In the foregoing solution, the first PE further includes: the CSF message sending module is used for sending CSF messages to the detection module; wherein,

the detection module is used for scanning the port state of the timed client access port, triggering the CSF message sending module when the port state is scanned as a port fault for the first time, and setting a CSF message sending identifier;

and the CSF message sending module is used for periodically sending the CSF message to the second PE after receiving the trigger information of the detection module.

In the above scheme, the detection module is further configured to clear the CSF message sending identifier and send a notification to the CSF message sending module to stop sending the CSF message after scanning that the port status is normal;

the CSF message sending module is further configured to stop sending the CSF message to the second PE after receiving the notification of stopping sending the CSF message sent by the detection module.

The scheme of the service convergence in the T-MPLS network provided by the invention comprises the steps that a PE connected with a CE scans the port state of a client access port of the PE at regular time, and sends a CSF message to the PE with a CE dual-homing function when the port state is a port fault; after the CE dual-homed PE receives the CSF message, updating the field value of whether the primary PW takes effect in the self-forwarding table entry to be invalid; and after receiving the data message, the CE dual-homed PE transmits the data message according to the updated forwarding table entry, so that when a link between the CE and the PE connected with the CE fails, the rapid convergence of the service can be realized.

In addition, the scanning interval of the PE connected with the CE for scanning the port state of the customer access port of the PE is not more than 10ms, so that the requirement that the carrier-class Ethernet service recovery is less than 50ms can be met, and the user experience is further improved.

Drawings

FIG. 1 is a network model of a typical CE dual homed PE;

FIG. 2 is a flow chart illustrating a method for service convergence in a T-MPLS network according to the present invention;

FIG. 3 is a schematic diagram of a T-MPPLS configuration according to an embodiment of the present invention;

fig. 4 is a schematic diagram of a system structure of service convergence in a T-MPLS network according to the present invention.

Detailed Description

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

In a T-MPLS network, a first CE accesses a second CE through a primary link, and in order to distinguish PE devices more clearly, in the following description, a PE directly connected to the second CE on the primary link is referred to as a PE connected to the CE, a PE doubly homed to the second CE at the other end on the primary link is referred to as a PE doubly homed to the CE, and the second CE is referred to as a remote CE.

The method for service convergence in a T-MPLS network of the present invention, as shown in fig. 2, includes the following steps:

step 201: the PE connected with the CE scans the port state of a client access port of the PE at regular time, and sends a CSF message to the PE with the CE dual-homing when the port state is a port fault;

here, the time interval of the timing scanning port state is set according to the user requirement, and when setting, in order to meet the condition that the time for recovering normal transmission of the service required by the current carrier ethernet is less than 50ms, the required time duration is not more than 10ms, and the user can arbitrarily set the timing time duration within 10ms, such as 0.5ms, 1ms, and the like.

The CSF message specifically includes: and notifying the CE of the failure of the dual-homed PE primary PW.

In the practical application process, the PE connected to the CE regularly scans the port state of its customer access port, and sends a CSF packet to the PE dual-homed to the CE when the port state is a port failure, specifically:

after receiving the trigger information of the detection module, the CSF message sending module periodically and automatically sends CSF messages to the PE to which the CE is dual-homed until receiving a notification of the detection module to stop sending CSF messages, and it is determined that CSF messages do not need to be sent any more. In the prior art, the cycle of sending the CSF message by the CSF message sending module is usually greater than the time interval of regularly scanning the port state, and generally 1 second is selected, so that when the detection module scans the port state again or is a port fault, the CSF message sending module does not need to be triggered again.

Here, the first time means that when the port state of the client access port is scanned last time before the scanning, the port state is port normal.

Before performing step 201, the method may further include:

setting T-MPLS tunnel head node information and example information on the CE dual-homed PE, setting T-MPLS tunnel tail node information and example information on the CE connected PE, forming a forwarding table item on the CE dual-homed PE, and then setting bidirectional forwarding detection on the CE dual-homed PE and the CE connected PE; the specific process of setting the head node information, the instance information and the tail node information of the T-MPLS tunnel is completely the same as the configuration process of the existing T-MPLS service, the process of forming the forwarding table entry is completely the same as the existing forming process, the detailed process of setting the bidirectional forwarding detection is specified in the TMPLS OAM protocol, and the specific process of setting the bidirectional forwarding detection is completely the same as the existing setting process.

The forwarding table entry comprises a tunnel table and an instance table; the tunnel table comprises public network labels, information of next hop of the tunnel, field values of whether the tunnel is effective or not, tunnel protection groups and other information; the instance table contains information such as a private network label, a main tunnel number corresponding to the private network label, an instance number, a field value indicating whether the PW is effective, a PW protection group and the like.

Step 202: after receiving the CSF message, the CE dual-homed PE updates a field value of whether the primary PW is in effect in its own forwarding entry to invalid, and then executes step 203;

here, when the field value of whether the active PW is valid indicates that the active PW is valid, the active PW can be used to send a data packet; and when the value of the field of whether the active PW takes effect is invalid, indicating that the active PW is invalid and not capable of sending data messages by using the active PW.

When bidirectional forwarding detection is simultaneously set on the CE dual-homed PE and the PE connected with the CE, when the PE connected with the CE fails, the CE dual-homed PE can also update the field value of whether the primary PW is in effect in the forwarding table entry to be invalid according to the detection result.

The CSF message sending module of the PE connected to the CE periodically and automatically sends the CSF message to the PE dual-homed to the CE before failing to receive the notification of the detection module to stop sending the CSF message, and when the PE dual-homed to the CE receives the CSF message again, the forwarding table entry is updated to the field value in which the primary PW fails, and therefore, after the PE dual-homed to the CE receives the CSF message again, the PE dual-homed to the CE checks whether the primary PW is the valid field value or the invalid field value in the forwarding table entry, and if the primary PW is the invalid field value, the forwarding table entry is not updated.

Step 203: and after receiving the data message which needs to be sent to the remote CE, the CE dual-homed PE sends the data message according to the updated forwarding table entry.

Specifically, the CE dual-homed PE queries an instance table in a forwarding table entry, and sends the data packet through the standby PW when it is found that the value of the field in effect of the primary PW is invalid. When sending, the PE of CE dual-homed inquires the tunnel table according to the tunnel number of the loading PW corresponding to the standby PW in the instance table, encapsulates the required information such as public network label and private network label into the data message according to the content corresponding to the tunnel table, and sends the encapsulated data message to the far-end CE through the standby PW.

If the current effective PW is the standby PW and the protection Mode of the PW protection group is a reversal Mode, when the main PW recovers to be normal and the port state of the PE connected with the CE is normal, after the time of WTR (Wait to Recovery) is passed, the PE of the CE double-homing updates the value of the field whether the main PW in the forwarding table is effective or not to be effective, and then sends a data message through the main PW; the method comprises the steps that a PE connected with a CE sends a Connectivity Check (CC) message to the PE connected with the CE through an active PW at regular time, if the PE connected with the CE can receive the CC message sent by the PE connected with the CE through the active PW within a specified time, the active PW is considered to be recovered to be normally used, and the specific processing process is completely the same as the existing processing process; and after the PE connected with the CE finds that the port state of the customer access port is normal through scanning, the PE does not send CSF messages to the CE dual-homed PE any more, and if the CE dual-homed PE does not receive the CSF messages within the specified time, the PE connected with the CE is considered to have the port state of normal.

The method of the present invention may further comprise: and the CE receives the data message sent by the PE with the CE dual homing.

The present invention will be described in further detail with reference to examples.

In this embodiment, based on the network model of the CE dual-homed PE shown in fig. 1, the application scenarios of this embodiment are as follows: PE _ A is a PE of CE _ B dual homing, PE _ B and PE _ C are both PEs connected with CE _ B, CE _ A accesses CE _ B, PE _ A is a head node of a T-MPLS tunnel, and PE _ B and PE _ C are tail nodes of the T-MPLS tunnel.

TUNNEL 1(TUNNEL1) and TUNNEL2 are respectively configured on PE _ A, PE _ B and PE _ C, specifically, TUNNEL1 is PE _ a-P _ a-PE _ B, TUNNEL2 is PE _ a-P _ B-PE _ C-PE _ B, TUNNEL1 is an active TUNNEL, TUNNEL2 is a standby TUNNEL, and an instance (SERVICE1) is configured on PE _ a, the SERVICE1 includes CIP1 and VIP1, wherein PW1 and PW2 are bound in VIP1, TUNNEL1 is carried by PW1, TUNNEL2 is carried by PW2, and PW1 and PW2 are configured to form active-standby protection, that is: PW1 is a primary PW, PW2 is a standby PW, and the configuration of the formed primary and standby paths is as follows:

TUNNEL1 carries PW1, corresponding main link: CE _ A-PE _ A-P _ A-PE _ B-CE _ B;

TUNNEL2 carries PW2, corresponding backup link: CE _ A-PE _ A-P _ B-PE _ C-CE _ B.

The configuration of PE _ B and PE _ C is the same as the configuration of PE _ a, and will not be described herein again, and when all nodes are configured, the configuration diagram shown in fig. 3 will be formed. After the configuration is completed, a forwarding table entry is formed on the PE _ a, which includes information of a public network tag, information of a next hop of the tunnel, a field value of whether the tunnel is in effect, a tunnel protection group, a private network tag, a main tunnel number corresponding to the private network tag, an instance number, a field value of whether the PW is in effect, a PW protection group, and the like.

And sets bi-directional forwarding detection on PE _ A, PE _ B.

PE _ B regularly scans the port state of a client access port of the PE _ B, and sends a CSF message to PE _ A when the port state is a port fault, and PE _ A PW1 is informed of failure; and after receiving the CSF message, PE _ A updates the field value of whether PW1 is valid in the self-forwarding table entry to be invalid.

When the PE _ B node fails, the PE _ a can know that the PE _ B node fails through the bidirectional forwarding detection result, that is: and knowing that the PW1 between the PE _ A and the PE _ B is invalid, and updating the value of the effective PW1 field in the self forwarding table entry to be invalid.

When the CE _ A has the data message sent to the CE _ B, after receiving the data message, the PE _ A queries an instance table in a self-forwarding table item, and if the PW1 is invalid, the PE _ A sends the data message through the PW 2. When the data message is sent through the PW2, PE _ a queries the TUNNEL table according to the PW-bearing TUNNEL number, i.e., TUNNEL2, corresponding to the PW2 in the instance table, and encapsulates the required information, such as the public network tag and the private network tag, in the data message according to the content corresponding to the TUNNEL table, that is: and encapsulating required information such as a private network label corresponding to PW2 and a public network label corresponding to TUNNEL2 into a data message, forwarding the encapsulated data message to PE _ C along a path between PE _ A and PE _ C, and forwarding the encapsulated data message to CE _ B according to a forwarding table entry on the PE _ C, so that services from CE _ A to CE _ B are forwarded, fast convergence of end-to-end services under link failure between PE _ B and CE _ B is realized, and meanwhile, the end-to-end service convergence time can be ensured to be less than 50 ms.

In order to implement the above method, the present invention further provides a system for service convergence in a T-MPLS network, as shown in fig. 4, the system includes: a first PE 41 and a second PE 42; wherein,

a first PE 41, configured to scan a port state of a client access port of the first PE at regular time, and send a CSF message to a second PE 42 when the port state is a port failure;

the second PE 42 is configured to update a field value indicating whether the primary PW is in effect in the self forwarding entry to be invalid after receiving the CSF message sent by the first PE 42; and after receiving the data message which needs to be sent to the remote CE, sending the data message according to the updated forwarding table entry.

The system may further include a setting unit 43, configured to set, on the second PE 42, T-MPLS tunnel head node information and instance information, set, on the first PE 41, T-MPLS tunnel tail node information and instance information, form a forwarding table entry on the second PE 42, and then set bidirectional forwarding detection on the first PE 41 and the second PE 42.

The first PE 41 may further include: the CSF message sending module is used for sending CSF messages to the detection module; wherein,

and a CSF message sending module, configured to send a CSF message to the second PE 42 periodically after receiving the trigger information of the detection module.

The second PE 42 is further configured to, when the active PW returns to normal and the port state of the PE connected to the CE is port normal, update the value of the field that whether the active PW is in effect to be in effect, and then send a data packet through the active PW.

The detection module is further configured to clear the CSF message sending identifier and send a notification to the CSF message sending module to stop sending CSF messages after the port status is scanned as normal;

the CSF message sending module is further configured to stop sending the CSF message to the second PE 42 after receiving the notification of stopping sending the CSF message sent by the detection module.

The system may further include a first CE configured to generate a data packet and send the data packet to the second PE 42.

The system may further include a second CE configured to receive the data packet sent by the second PE 42.

The above description is only exemplary of the present invention and should not be taken as limiting the scope of the present invention, and any modifications, equivalents, improvements, etc. that are within the spirit and principle of the present invention should be included in the present invention.

Claims

1. A method for traffic convergence in a transport multi-protocol label switching (T-MPLS) network, the method comprising:

the provider edge equipment (PE) connected with the user side edge equipment (CE) scans the port state of the client access port of the PE at regular time, and sends a Client Signal Failure (CSF) message to the PE of the CE dual-homing when the port state is a port failure;

2. The method according to claim 1, wherein the timing duration of the PE connected to the CE scanning the port status of its own customer access port is not more than 10 ms.

3. The method according to claim 1 or 2 or the above, wherein before said PE connected to the CE periodically scans the port status of its own customer access port, the method further comprises:

4. The method according to claim 3, wherein the sending the data packet according to the updated forwarding table entry is:

5. The method of claim 4, further comprising:

6. The method according to claim 1 or 2, wherein the PE connected to the CE periodically scans the port status of its customer access port, and when the port status is port failure, sends a CSF message to the PE dual-homed to the CE, where the CSF message is:

7. The method of claim 6, further comprising:

8. A system for traffic convergence in a T-MPLS network, the system comprising: a first PE and a second PE; wherein,

9. The system of claim 8, further comprising a setting unit configured to set T-MPLS tunnel header node information and instance information on the second PE, set T-MPLS tunnel tail node information and instance information on the first PE, form a forwarding table entry on the second PE, and then set bi-directional forwarding detection on the first PE and the second PE.

10. The system according to claim 8 or 9, wherein the second PE is further configured to update the value of the effective-or-not field of the active PW to be effective when the active PW recovers to normal and the port status of the PE connected to the CE is port normal.

11. The system according to claim 8 or 9, wherein the first PE further comprises: the CSF message sending module is used for sending CSF messages to the detection module; wherein,

12. The system of claim 11,