CN1601965A - Route path fault repairing method - Google Patents

Abstract

The invention discloses a method for repairing a routing path fault, comprising: a resource manager of a bearer network searches for an equivalent path of the current abnormal label switching path for a session, searches for a resource request of the session, and converts the current abnormal label switching path in the service routing path information Replace it with an equivalent path, and send a resource change request to the upstream bearer network resource manager until the request is sent to the source bearer network resource manager of the session; the source bearer network resource manager of the session receives the resource change request message, Update the resource allocation information, and issue a flow mapping command to the ingress edge router of the session, and use the updated routing path to forward the service flow. This solution can quickly rebuild the routing path when the LSP state changes abnormally, so that business services can continue.

Description

A Routing Path Fault Restoration Method

technical field

The invention relates to the technical field of network communication, in particular to a method for repairing a bearer layer routing path failure.

Background technique

With the continuous increase of Internet scale, various network services emerge one after another, and multimedia systems including voice, image and other services emerge in endlessly. Since real-time services are sensitive to characteristics such as network transmission delay and delay jitter, when there are services such as file transfer (FTP) or hypertext transfer (HTTP) containing image files on the network, the real-time service will fail. On the other hand, multimedia services take up a large amount of bandwidth, so it is difficult for the key services to be guaranteed by the existing network to be reliably transmitted.

As a result, various Quality of Service (QoS, Quality of Service) technologies emerged as the times require. To meet the requirements of QoS, the Internet Engineering Task Force (IETF, Internet Engineering Task Force) has proposed many service models and mechanisms. At present, the industry generally recognizes that the integrated service model (Int-Serv) is used at the access and edge of the network, and the differentiated service model (Diff-Serv) is used at the core of the network. However, Diff-Serv only sets priority levels to ensure QoS measures. Although the line utilization rate is improved, the specific transmission effect is difficult to predict. Therefore, the industry began to introduce an independent bearer control layer for the backbone network differentiated service Diff-Serv, and established a set of dedicated Diff-Serv QoS signaling mechanisms to manage network topology resources at the bearer control layer, collectively referred to as this resource management differentiated service Diff -Serv mode is a Diff-Serv model with an independent bearer control layer.

Refer to Figure 1 for the topological structure of the Diff-Serv network model with an independent bearer control layer. Above the bearer layer where the bearer network is located, the bearer control layer and the service control layer are further divided. The call agent (CA, Call Agent) 102 of the service control layer is used to act as an agent for the user's call, and forward the call to the bearer control layer. There are multiple bearer network resource managers 101 in the bearer control layer, and the range managed by each bearer network resource manager 101 in the bearer layer is called the management domain 103 of the bearer network resource manager 101, and each management domain 103 There are several routers connected to the bearer network resource manager 101, and the circles shown in the figure represent routers. Usually, management rules and network topology are configured in the bearer network resource manager 101, which can apply for user services, establish paths, allocate resources, and issue management control commands to bearer layer routers. The bearer network resource managers 101 of each management domain 103 can also transmit the client's service bandwidth application request and result, and the path information allocated by the bearer network resource managers 101 for the service application, etc. through signaling.

When establishing a session of services such as voice, the user (UE) sends a connection request to CA 102, and CA 102 sends the connection request to the corresponding bearer network resource manager 101, and the bearer network resource manager 101 analyzes the connection request and provides Allocate the corresponding resources for the connection, select the appropriate routing path, send the policy routing information to the corresponding router in the domain, and establish the session. Each label switched path (LSP) in the bearer network bears at least one session, which is distinguished by a specific session identifier, and each session generally sends data through multiple service flows. If an LSP is disconnected, all service flows cannot be sent. At this time, the router related to the LSP can be made to report the change of the LSP state to the resource manager of the bearer network. When the bearer network resource manager receives the LSP status change, it usually deletes all calls carried on this LSP, and notifies the CA to delete related content. But the session will be interrupted like this, and the user must call again if he wants to continue the call, which brings inconvenience to the use of the user, causes waste of resources, and affects the development of network services.

Contents of the invention

In view of this, the main purpose of the present invention is to provide a routing path failure repair method, when a certain section of the LSP state on the routing path changes abnormally, the routing path and the session routing path on it are quickly repaired, so that business services can continue.

A method for repairing routing path failures, comprising the following steps:

a) The bearer network resource manager belonging to the management domain where the abnormal label switching path is located searches for the equivalent path of the current abnormal label switching path for the session;

b) The current bearer network resource manager searches for the resource request of the session, and replaces the current abnormal label switching path in the service routing path information of the session with an equivalent path, if the current bearer network resource manager is not the source bearer network resource of the session manager, then the current bearer network resource manager sends a resource change request message to change the current abnormal label switching path to an equivalent path to the previous hop bearer network resource manager along the session routing path, and the last hop bearer network resource manager After receiving the message, the device joins the path information in its own management domain and continues to forward upstream until the source bearer network resource manager of the session, and enters step c), if the current bearer network resource manager is the source bearer network resource manager of the session resource manager, enter step c);

c) The source bearer network resource manager of the session updates the resource allocation information according to the received resource change request message, and sends a flow mapping command to the ingress edge router of the session, and the bearer layer of the network uses the updated routing path to forward the session business flow.

2. The method according to claim 1, characterized in that, after step c), further comprising:

d1) Judging whether there are unrepaired sessions on the current abnormal label switching path, if yes, return to step a), otherwise, end the process.

3. The method according to claim 1, wherein the equivalent path in the method is a combination of label switching paths in another or more than one domain that replaces the routing path of the current session in the management domain where the current abnormal label switching path resides, or replaces A combination of intra-domain and inter-domain label-switched paths for the current session routing path.

4. The method according to claim 1, characterized in that, after step c), further comprising:

d2) Judging whether there are unrepaired sessions on the current abnormal label switching path, if so, directly adopt the equivalent path found in step a), and return to step b), otherwise, end the process.

5. The method according to claim 4, wherein said equivalent path is a combination of another or more than one label switching path which has a common egress router and ingress router with the current abnormal label switching path.

6. The method according to claim 1, characterized in that, after sending the flow mapping command to the ingress edge router in step c), further comprising:

c1) Judging whether the routing path reconstruction of the current session is successful, if so, the bearer layer of the network uses the updated routing path to forward the service flow of the session, otherwise, delete the session and release the session resource, and end the process.

7. The method according to claim 1, wherein step a) further comprises: judging whether the equivalent path is searched successfully, if yes, proceed to step b), otherwise, delete the session and release the session resource, and end the process .

8. The method according to claim 1, wherein, if the bearer network resource manager finds more than one equivalent path of the current abnormal label switched path in step a), further according to load sharing or service type, resource availability , priority, and some specific QoS requirements, select one of the more than one equivalent paths found through the local path selection strategy.

9. The method according to claim 1, wherein in step a), if the management domain where the current abnormal label switching path is located is jointly managed by more than one bearer network resource manager, then the current abnormal label is searched in one bearer network resource manager After the equivalent path of the switched path fails, another bearer network resource manager of the management domain is assigned to continue searching for the equivalent path of the current abnormal label switched path.

10. The method according to claim 1, wherein in step c), the resource allocation information update content of the session at least includes: label stack information of the session routing path.

11. The method according to claim 1, wherein the rebuilding command in step c) at least includes: current session information, QoS parameters, and label stack flow mapping command notification of the session routing path.

12. The method according to claim 1, wherein the source bearer network resource manager is the home bearer network resource manager of the calling end office router or tandem office router of the current session, or the called end office router Or the home bearer network resource manager of the gateway router.

13. The method according to claim 1, further comprising: setting the state of the current abnormal label switched path as abnormal, then in step a) the resource manager of the bearer network first judges whether the state of the label switched path is Abnormal, if yes, do nothing, end the process, otherwise, continue the process;

And in the process of finding the equivalent path, it is further judged whether the found equivalent path includes a label switched path with an abnormal state, if yes, the equivalent path is excluded, otherwise, the equivalent path is accepted.

14. The method according to claim 1, further comprising: setting the number or time of equivalent path reconstruction for the same abnormal label switched path, step a) the resource manager of the bearer network first judges whether the abnormal label switched path Exceed the set rebuild times or time, if yes, end the process, otherwise, continue the process.

As can be seen from the above scheme, the routing path failure repairing method of the present invention, when the status of a section of the LSP on the routing path of the session changes abnormally, by searching for the LSP equivalent path and updating the resource allocation information, the routing path is quickly realized. Dynamic repair without interrupting the current business of users, avoiding the waste of network resources, improving the success rate of business services, and enhancing the stability of network operations, and the invention is applicable to networks of any scale, and has the advantages of simple implementation and easy network Status maintenance management and other advantages.

Description of drawings

Fig. 1 is a schematic diagram of the topological structure of a Diff-Serv network model with an independent bearer control layer;

FIG. 2 is a schematic diagram of a network structure using dual planes for resource management;

Fig. 3 is a flowchart of an embodiment of the present invention.

Detailed ways

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

The main idea of the present invention is that after the bearer network resource manager receives an LSP state abnormality message in its management domain, it searches for the equivalent path of the current abnormal label switching path for the session, searches for the resource request of the session, and converts the service routing path information in the domain Replace the current abnormal label switching path with an equivalent path, and send a resource change request to the upstream bearer network resource manager until the request is sent to the source bearer network resource manager of the session; the source bearer network resource manager of the session receives resource change request message, update the resource allocation information, and issue a flow mapping command to the ingress edge router of the session, and use the updated routing path to forward the service flow. In this way, until all the sessions carried on the current abnormal LSP are all repaired.

Referring to FIG. 2, FIG. 2 is a schematic diagram of a network structure using dual planes for resource management. In this embodiment, a Connection Manager (CM, Connection Manager) 201 is used as the bearer network resource manager. In this network structure, the same management domain can be jointly managed by two CMs 201, and each CM 201 manages a part of routers in the domain, such as CM2 and CM3 in the figure. The black circle in the figure represents the edge router (ER, Edge Router), and the gray circle represents the border router (BR, Border Router). Assuming that a certain LSP in the CM1 management domain in the network fails, if there is a session through this segment of LSP, it will be detected This LSP state changes abnormally, and is notified to CM1, and the routing path fault repair process of the present invention is followed, referring to the flow chart shown in Figure 3:

Step 301, after CM1 receives the LSP state abnormal change message in its management domain, it searches for the equivalent path of the LSP in its management domain, if it finds an equivalent path, then allocates the corresponding bandwidth for the current session, and enters the next step, otherwise, If the rerouting process fails, go to step 304.

Wherein, the equivalent path can also be any intra-domain routing path that can replace the routing path of the current session in the management domain where the current abnormal LSP is located, that is, the session can replace the routing path of the session between the ingress router and the egress router of the management domain For example, the routing path of a certain session in Figure 2 needs to pass through LSP1+LSP2. If the state of LSP1 is abnormal, the path of LSP3+LSP4 can be selected. The equivalent path may also be another LSP that has a common egress router and ingress router with the current abnormal LSP, or a combination of multiple LSPs. For example: in Figure 2, if CM1 receives the abnormal state change message of LSP1, the equivalent path of LSP1 can be LSP3+LSP5; similarly, if the state of LSP1+LSP2 changes, its equivalent path can be LSP3+LSP4. If there are multiple equivalent paths, you can select one of them through local configuration path selection policy according to load sharing or service type, resource availability, priority, and some specific QoS requirements.

The LSP state change message that CM1 described in this step receives can be from the connection node (CN, Connection Node) in its domain, and wherein, CN refers to ER or BR or core router, and, LSP state change message also can be from obtained by other means.

In step 302, because one LSP can bear multiple sessions, CM1 may be the source CM of the session or a CM in other locations. If CM1 is the source CM of the session, CM1 will search for the resource request of the current session on the abnormal LSP in the current state in its saved session resource requests, and replace the current abnormal label switching path in the service routing path information of the session in the domain with the equivalent path, go to step 303.

If CM1 is not the source CM of the session, CM1 searches for the resource request of the current session in its own saved information, replaces the current abnormal label switching path in the service routing path information of the session in the domain with an equivalent path, and then routes along the session The path sends a resource change request message to the last hop CM of this CM, notifying to change the routing path segment corresponding to the abnormal LSP on the routing path of the current session to an equivalent path. After the last hop CM receives the message, plus the The service routing path information continues to forward the resource change request message upstream until the source CM of the session.

The source CM described in the present invention is the attributable CM of the calling end office router or the calling tandem office router of the current session, and may also refer to the attributable CM of the called end office router or the tandem office router. The forward resource request is Established along the direction from the calling party to the called party, then the source CM refers to the home CM of the calling end office router or the calling tandem office router; the reverse resource request is established along the direction from the called party to the calling party, then The source CM refers to the home CM of the called end office router or the called tandem office router. If two-way binding establishes a resource request, the source CM can be any one of the home CMs of the calling/called end office router or the tandem office router.

Step 303, if CM1 is the source CM of the session, CM1 updates the resource allocation information according to the received resource change request message, and the updated resource allocation information mainly includes: label stack information of the session routing path. Then, issue a flow mapping command to the ingress ER of the session in its management domain, and notify the bearer layer of the new routing path through the ER, so that the bearer layer uses the new routing path to forward the service flow of the session. Wherein, the flow mapping command includes at least current session information, QoS parameters and a label stack of the session routing path.

If CM1 is not the source CM of the session, the source CM of the session performs the same operation process as that of CM1 in step 303 after receiving the resource change request message forwarded downstream.

The specific process of updating the routing path in the bearer layer has been specified in the Label Distribution Protocol (LDP) and the Resource Reservation Protocol (RSVP) and other protocols. According to the different needs of the network, the teardown of the connection and the establishment and distribution of labels can be adopted. Many different processing methods.

Step 304, judging whether the reconstruction is successful, if the reconstruction is successful, then enter the next step, if the routing path reconstruction of the session fails due to the exhaustion of resources or other reasons, then CM1, that is, the source CM of the session, deletes the session and releases the associated Resource, the source CM of the session reports the failure message containing the session information and reason code to the CA, and also sends a resource release request to all CMs on the entire routing path to release the session resource.

Step 305, judge whether there are other unreestablished sessions on the LSP, if so, return to step 301, continue to search for an equivalent path for the next session related to the LSP, and rebuild the session routing path, otherwise, if not, end process. In this way, until all sessions on the LSP are reestablished.

In addition, in the process of rebuilding the session routing path on the same abnormal LSP, in order to avoid returning to step 301 to search for an equivalent path every time the present invention rebuilds, it is also possible to specify that the selected equivalent path only It can be another LSP that has a common egress router and ingress router with the current abnormal LSP, or a combination of multiple LSPs. In this way, since the selected equivalent path can completely replace the current abnormal LSP, when repairing the routing path of the next session , may return to step 302, and directly perform the operation of reselecting the session routing path. At this time, the repair speed of the route is faster, but it is not necessarily the optimal path for each session, and since the bandwidth to be allocated for each session may be different, it is necessary to consider when allocating the bandwidth so that the bandwidth can satisfy all Session requirements.

Step 306, CM1, that is, the source CM, deletes the session and releases the relevant resources of the session, and the source CM of the session reports a failure message containing session information and reason codes to CA, and also sends a resource release request to all CMs on the entire routing path, Release the session resource.

In addition, in a network using dual planes for resource management as shown in Figure 2, or a network for multi-plane resource management, since the routers in one management domain may be managed by more than one CM respectively, if an LSP status change is received If the CM of the message has not found the equivalent path of the abnormal LSP in the router it manages, it can also send a resource change request message to other CMs belonging to the domain, and the other CMs belonging to the domain will search for the equivalent LSP path it manages , if found, then enter step 302 to perform the following operations, if not found, the resource change request message can be forwarded to other CMs belonging to the domain to continue searching, if all CMs belonging to the domain cannot be rebuilt, then end the process. For example, as shown in FIG. 2 , if the status of a certain LSP in CM2 changes, CM2 can ask CM3 to continue searching for the equivalent route of the LSP, and then enter step 302 .

Furthermore, if an equivalent intra-domain routing path cannot be found, the inter-domain LSP between the CM and its upstream and downstream CMs can be changed first, and an intra-domain routing path is reselected in the CM, that is, the equivalent path is extended to the inter-domain.

In addition, for an LSP that is in an abnormal state, because its relevant information has not been updated, the routing path passing through the LSP may still be selected during the routing path selection or equivalent path selection process of a new session call. If the LSP is still in In the abnormal state, the rerouting process still needs to be carried out. Therefore, in order to prevent this situation from happening again, you can set the status flag for the LSP after the LSP equivalent path is rebuilt. During the routing path or equivalent path selection process of the session, By checking the state flag of the LSP, it is known that the state of the LSP is abnormal, and other routing paths can be selected. And in step a), the CM first judges whether the LSP is in an abnormal state, if so, does not do any processing, and ends the process; otherwise, continues the process. And in the process of finding the equivalent path, it is further judged whether the found equivalent path includes a label switched path with an abnormal state flag, if yes, the equivalent path is excluded, otherwise, the equivalent path is accepted.

In addition, you can also set the number of rerouting for the same LSP failure. If the routing path failures reported for multiple times are all for the same LSP and exceed the preset times of reconstruction and repair, the session routing or equivalent path When selecting, the routing path passing through the LSP is no longer selected, and the routing path of the LSP is no longer repaired. And in step a) the bearer network resource manager first judges whether the abnormal label switched path exceeds the set reconstruction times or time, if yes, then end the process, otherwise, continue the process. Similarly, timers can also be used to complete the above functions. For example, the CM can refuse to select this LSP if it still receives the same LSP status change message after repairing the routing path several times within a certain period of time.

The above descriptions are only preferred embodiments of the present invention, and are not intended to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present invention shall be included in the scope of the present invention. within the scope of protection.

Claims (14)

1. A routing path failure repair method, characterized in that, comprising the following steps: a) The bearer network resource manager belonging to the management domain where the abnormal label switching path is located searches for the equivalent path of the current abnormal label switching path for the session; b) The current bearer network resource manager searches for the resource request of the session, and replaces the current abnormal label switching path in the service routing path information of the session with an equivalent path, if the current bearer network resource manager is not the source bearer network resource of the session manager, then the current bearer network resource manager sends a resource change request message to change the current abnormal label switching path to an equivalent path to the previous hop bearer network resource manager along the session routing path, and the last hop bearer network resource manager After receiving the message, the device joins the path information in its own management domain and continues to forward upstream until the source bearer network resource manager of the session, and enters step c), if the current bearer network resource manager is the source bearer network resource manager of the session resource manager, enter step c); c) The source bearer network resource manager of the session updates the resource allocation information according to the received resource change request message, and sends a flow mapping command to the ingress edge router of the session, and the bearer layer of the network uses the updated routing path to forward the session business flow. 2. The method according to claim 1, characterized in that, after step c), further comprising: d1) Judging whether there are unrepaired sessions on the current abnormal label switching path, if yes, return to step a), otherwise, end the process. 3. The method according to claim 1, wherein the equivalent path in the method is a combination of label switching paths in another or more than one domain that replaces the routing path of the current session in the management domain where the current abnormal label switching path resides, or replaces A combination of intra-domain and inter-domain label-switched paths for the current session routing path. 4. The method according to claim 1, characterized in that, after step c), further comprising: d2) Judging whether there are unrepaired sessions on the current abnormal label switching path, if so, directly adopt the equivalent path found in step a), and return to step b), otherwise, end the process. 5. The method according to claim 4, wherein said equivalent path is a combination of another or more than one label switching path which has a common egress router and ingress router with the current abnormal label switching path. 6. The method according to claim 1, characterized in that, after sending the flow mapping command to the ingress edge router in step c), further comprising: c1) Judging whether the routing path reconstruction of the current session is successful, if so, the bearer layer of the network uses the updated routing path to forward the service flow of the session, otherwise, delete the session and release the session resource, and end the process. 7. The method according to claim 1, wherein step a) further comprises: judging whether the equivalent path is searched successfully, if yes, proceed to step b), otherwise, delete the session and release the session resource, and end the process . 8. The method according to claim 1, wherein, if the bearer network resource manager finds more than one equivalent path of the current abnormal label switched path in step a), further according to load sharing or service type, resource availability , priority, and some specific QoS requirements, select one of the more than one equivalent paths found through the local path selection strategy. 9. The method according to claim 1, wherein in step a), if the management domain where the current abnormal label switching path is located is jointly managed by more than one bearer network resource manager, then the current abnormal label is searched in one bearer network resource manager After the equivalent path of the switched path fails, another bearer network resource manager of the management domain is assigned to continue searching for the equivalent path of the current abnormal label switched path. 10. The method according to claim 1, wherein in step c), the resource allocation information update content of the session at least includes: label stack information of the session routing path. 11. The method according to claim 1, wherein the rebuilding command in step c) at least includes: current session information, QoS parameters, and label stack flow mapping command notification of the session routing path. 12. The method according to claim 1, wherein the source bearer network resource manager is the home bearer network resource manager of the calling end office router or tandem office router of the current session, or the called end office router Or the home bearer network resource manager of the gateway router. 13. The method according to claim 1, further comprising: setting the state of the current abnormal label switched path as abnormal, then in step a) the resource manager of the bearer network first judges whether the state of the label switched path is Abnormal, if yes, do nothing, end the process, otherwise, continue the process; And in the process of finding the equivalent path, it is further judged whether the found equivalent path includes a label switched path with an abnormal state, if yes, the equivalent path is excluded, otherwise, the equivalent path is accepted. 14. The method according to claim 1, further comprising: setting the number or time of equivalent path reconstruction for the same abnormal label switched path, step a) the resource manager of the bearer network first judges whether the abnormal label switched path Exceed the set rebuild times or time, if yes, end the process, otherwise, continue the process.