JP2000278305A - Fault evading method for router and router node - Google Patents

Abstract

PROBLEM TO BE SOLVED: To evade a faulty node by declaring the soundness of the faulty node which is disconnected, and validating an internal routing table of the faulty node and implementing the function of the faulty node instead. SOLUTION: If a fault occurs, the node B is disconnected from a loop network and a network excluding the node B is reconstituted. At the nodes A and B at both the ends of the faulty node B, the routing table of the node B and its processing process (b) are validated in addition to processing processes (a) and (c). Consequently, the processing process (b) present at the node A behaves for the processing process (a) and the processing process (b) present at the node C behaves for the processing process (c) as if the node B were in normal operation. Therefore the nodes A and B carry out the routing process of a network layer that the node B originally should perform.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a router which constitutes a node of a ring-shaped redundant network and has a routing table, and a method for avoiding a failure of a router node.

[0002]

2. Description of the Related Art As a measure against network failure, for example, in SONET, a ring switch (BLSR: BLSR) is used in which a network is duplicated in a ring shape, and a failure transmission path or a failure node is separated from each node to reconfigure a ring network. bi-directional line-switched ri
ng) is adopted.

[0003] The nodes that make up the SONET ring network are transmission / reception devices, but since the failure avoidance by the ring switch is completed within about 50 milliseconds, it is possible to provide a very high-performance protection function. It is possible.

[0004] Therefore, a routing network having a router as a node is a backbone network that requires high reliability. Therefore, similarly, as shown in FIG. If a switch having the function of the above-described ring switch is provided in each of the routers, the faulty node including the transmission line fault can be separated and the ring network can be reconfigured as shown in FIG.

Hereinafter, a specific description will be given with reference to FIGS. FIG. 3 shows a topology in a normal state, and FIG. 4 shows a topology in a state where a failed node is avoided. In the following description, “layer” refers to the OS
The terms “network layer” and “data link layer” in the hierarchical structure of the I reference model are used.

In FIG. 3, this routing network is composed of nodes E, F, G, and H.
The transmission line is duplicated with a working line W of the active system and a protection line P of the standby system. Each node includes the switch means described above.

FIG. 3 shows a normal state in which the nodes E and G are performing communication via the working line W via the node F. The routing network layer recognizes the topology of the data link layer in this normal state as a ring (square) composed of four nodes.

[0008] For example, when a failure occurs in the node F, the switch means operates in the adjacent nodes E and G, and the node F is disconnected from the network.
The ring network using the protection line P is reconfigured.

As a result, the node E and the node G shift to communication via the node H by the protection line P.
The topology of the data link layer after the failure avoidance is recognized as a ring (triangle) composed of three nodes.

[0010]

However, FIG. 3 and FIG.
In a configuration in which a failed node is simply separated from the network as shown in Fig. 7, the network layer performing routing recognizes a change in the topology of the data link layer. Communication occurs, and the reconstruction of the routing table starts. As a result, it takes several seconds to recover the network, which makes it difficult to ensure reliability.

An object of the present invention is to provide a router having means for avoiding a failed node by making a topology change only by a data link layer without being conscious of an upper network layer, and a method for avoiding a failure of a router node. It is in.

[0012]

According to a first aspect of the present invention, there is provided a router which comprises a node of a ring-shaped redundant network and has a routing table of its own node, wherein each router has a routing table of two adjacent nodes. When a failure of an adjacent node is detected, the failure node is disconnected from the network, and a switch means for reconfiguring the ring network using the backup transmission line is used. In addition to declaring that the failed node is alive, the device includes means for validating a built-in routing table of the failed node and acting as a substitute for the function of the failed node.

According to the first aspect of the present invention, when a node failure including a transmission line failure occurs, the nodes located on both sides of the failed node reconfigure the ring network excluding the failed node. At the same time, the function of the failed node is performed on behalf of the failed node, and a state as if the failed node does not exist is simulated.

Therefore, even when a node failure occurs and the topology is changed in the data link layer, the change can be made invisible to the upper layer.

According to a second aspect of the present invention, there is provided the router according to the first aspect, wherein the routing information of the own node operated by the own node is transmitted to the active transmission line, and is fetched from the active transmission line. Means for sending routing information to the protection transmission line, means for updating the routing table of the own node by the routing information sent via the working transmission line, and routing of the own node operated in the own node. Means for updating a routing table for an adjacent node incorporated in the own node based on the information and the routing information fetched from the backup transmission line.

According to the second aspect of the present invention, in each node, in a normal state, the routing table can be updated using the backup transmission line. Therefore, effective use of the standby transmission line can be achieved.

According to a third aspect of the present invention, in the router according to any one of the first and second aspects, when the router is located on both sides of the separated failed node, the built-in failed node A routing table is updated with routing information based on the contents of the routing table of the own node, and a means for transmitting the routing information to an adjacent node is provided.

According to the third aspect of the present invention, in a node acting as a substitute for a failed node, the routing table of the failed node is updated. Therefore,
Routing can be performed without trouble in a reconfigured network excluding a failed node.

According to a fourth aspect of the present invention, there is provided a router node failure avoiding method, wherein each router constituting a node of a ring-shaped redundant network includes routing information of two adjacent nodes in addition to routing information of its own node. When a node failure occurs, the nodes located on both sides of the failed node separate the failed node from the network, reconfigure the ring network, and use the built-in routing information of the failed node to perform the reconfiguration. It is characterized in that the function of the failed node is substituted.

According to a fifth aspect of the present invention, there is provided the router node failure avoiding method according to the fourth aspect, wherein the routing information of the own node operated by the own node is transmitted to the working transmission line. Each node updates the routing table of its own node and the routing table related to the adjacent node incorporated in its own node by transmitting the routing information fetched from the working transmission line to the backup transmission line. It is characterized by the following.

According to a sixth aspect of the present invention, there is provided the router node failure avoiding method according to any one of the fourth and fifth aspects, wherein the router node is located on both sides of the separated failed node. The node router is
It is characterized in that the built-in routing information of the failed node is updated based on the routing information of the own node, and the routing information is transmitted to the adjacent node facing the node.

[0022]

Embodiments of the present invention will be described below with reference to the drawings.

FIG. 1 shows an example of the configuration of a routing network having a ring-shaped duplex configuration configured using a router according to an embodiment of the present invention. This embodiment corresponds to claims 1 to 6. However, FIG. 1 mainly shows a management state of the routing table of each node in a normal state.

In FIG. 1, the routing network includes four nodes A, B, C, and D.
The transmission line is duplicated with a working line W of the active system and a protection line P of the standby system.

The working line W and the protection line P are an optical fiber transmission line or a metallic transmission line, respectively.

The routers constituting each node are IP
Although it is a router or a multi-protocol router, in addition to the routing table of its own node and its processing process, it has a routing table of both adjacent nodes and its processing process.

Specifically, the node A includes the routing tables of the adjacent nodes B and D and the processing processes b and d in addition to the routing table of the own node and the processing process a thereof. The node B includes the routing tables of the adjacent nodes A and C and the processing processes a and c thereof in addition to the routing table of the own node and the processing process b thereof.

Similarly, the node C includes the routing tables of the adjacent nodes B and D and the processing processes b and d in addition to the routing table of the node itself and the processing process c. The node D includes the routing tables of both adjacent nodes A and C and the processing processes a and c in addition to the routing table of the own node and the processing process d.

Although not shown, each node transmits / receives routing information to / from an adjacent node via the working line W and the protection line P, and separates the faulty node from the network, including a fault in the transmission path, and disconnects the faulty node from the network. Switch means for reconfiguring the ring network using P.

In this switch means, the transmission system is an SDH network (S
(Synchronous Digital Hierarchy)
An ONET ring switch (BLSR) can be applied. In ITU-T, this BLSR is referred to as MSSPR (Mult
iplex Section Shared Protection Ring).

Next, the operation of this embodiment will be described with reference to FIGS. FIG. 2 is an explanatory diagram of the operation of avoiding a failed node according to the present embodiment. However, FIG. 2 mainly shows the management state of the routing table of each node after avoiding the failed node.

First, in each node, the routing table is updated in the normal state as follows. In FIG. 1, the update of the routing table of the own node is performed by an update message sent from the adjacent node on the working line W.

On the other hand, the updating of the routing table for both adjacent nodes is performed by two methods. 1
First, as indicated by a broken line in FIG. 1, an update message issued from the own node to each adjacent node is directly received internally. For example, node D
In (B), an update message for node C is sent to the working line W, which is sent directly to node D.
This is taken into the routing table for node C provided in (B).

The other is that at each adjacent node, the update message received by the adjacent node from the working line W is received via the protection line P from the adjacent node. For example, node D
The update message for node C sent to the working line W by (B) is sent to node B (D) at node C.
Therefore, the node D (B) fetches the update message for the node C from the protection line P.

In the other nodes A and C, the routing table is updated for the adjacent nodes on both sides by the same operation.

Next, the operation of avoiding a fault will be described. Figure 1,
In FIG. 2, for example, when a failure occurs in the node B, the switching means of the nodes A and C detect no signal and start up, respectively, and the nodes A and C separate the node B from the loop network and protect the line. A loopback operation using P is performed to reconfigure the ring network excluding the node B. as a result,
The loop network transitions from the state of FIG. 1 to the state of FIG. The transition time is approximately 50 milliseconds if the switching means is a ring switch (BLSR).

The processing process b existing in the nodes A and C sends keep-alive messages to the processing processes a and c, respectively, in response to the activation of the switch means of the nodes A and C. And informs that node B is alive.

Thus, as shown in FIG. 2, the nodes A and C at both ends of the failure node B have their own routing tables and their processing processes a and c.
In addition to Node B, which was updated in the normal state
Of the routing table and the processing process b thereof become effective. In FIG. 2, since the node D is not involved in the current failure avoidance, only the routing table of its own node and its processing process d are shown as if they are valid.

As a result, it is as if the processing process b existing at the node A is for the processing process a, the processing process b existing at the node C is for the processing process c, and the node B is operating normally. Node A
And the node C will take over the routing process at the network layer that should have been performed at the node B.

Then, in the nodes A and C that have performed the failure avoidance, the routing table of the failed node B is updated as follows. That is, when the processing process b existing in the node A internally receives the update message from the processing process a, the processing process b updates the routing table of the processing process b, and further, the update message received from the processing process a is present in the node C. To the processing process b via the protection line P.

When receiving the update message from the process b of the node A via the protection line P, the processing process b of the node C updates the routing table for the node C, and if necessary, reflects the result. And internally sends an update message to the processing process c. The same applies to the propagation of the update message from the node C to the node A.

As described above, according to the present embodiment, a change in the network topology due to the avoidance of a failure can be absorbed by the data link layer and made invisible to the network layer. Reconstruction of the routing table does not occur, and closed protection can be performed only on the data link layer. When the transmission system is an SDH network, the data link layer is at an MS (Multiplex Section) level.

[0043]

As described above, according to the first and fourth aspects of the present invention, when a node failure including a transmission line failure occurs in a duplicated loop network, the nodes located on both sides of the failed node are removed. In this case, a state in which the failed node does not exist is simulated. Therefore, even if a node failure occurs and the topology is changed in the data link layer, it is possible to make the upper layer (network layer) performing the routing recognize as if the network has not changed and is still alive. Therefore, the connection state is not changed or the routing table is not reconstructed in the upper layer, and the restoration can be quickly performed by the processing of only the data link layer, and high-performance node failure avoidance can be performed.

According to the second and fifth aspects of the present invention, in each node, in the normal state, the routing table can be updated using the backup transmission line, so that the routing table can be effectively used while the backup transmission line is effectively used. Can always be kept up-to-date to prepare for a failure.

According to the third and sixth aspects of the present invention, the routing table of the failed node can be updated in the node acting as the failed node, so that there is no problem in the reconfigured network excluding the failed node. Routing is possible.

[Brief description of the drawings]

FIG. 1 is a configuration diagram (mainly showing a management state of a routing table) of a routing network having a ring-shaped duplex configuration configured using a router according to an embodiment of the present invention.

FIG. 2 is an explanatory diagram (mainly showing a management state of a routing table) of an operation of avoiding a failed node according to the embodiment;

FIG. 3 is a diagram showing a topology in a normal state of a routing network having a ring-shaped duplex configuration.

FIG. 4 is a diagram showing a topology in a state where a failed node in a routing network having a ring-shaped duplex configuration is avoided.

[Explanation of symbols]

A, B, C, D Node W Working line P Protection line a Routing table of node A and its processing process b Routing table of node B and its processing process c Routing table of node C and its processing process d Routing table of node D And its processing

Claims (6)

[Claims] When each router that configures a node of a ring-shaped redundant network and has a routing table of its own node detects a failure of a neighboring node and a failure of the neighboring node, the failure node is removed from the network. A switch means for reconfiguring a ring network using a detached backup transmission line, and a switch means located on both sides of the detached failed node,
A router comprising: means for declaring that the separated failed node is alive and validating a built-in routing table of the failed node and acting on behalf of the function of the failed node. 2. The router according to claim 1, wherein the routing information of the own node operated by the own node is transmitted to an active transmission line, and the routing information fetched from the active transmission line is transmitted to a standby transmission line. Means for updating the routing table of the own node with the routing information sent via the working transmission line, and the routing information of the own node operated in the own node and fetched from the backup transmission line. Means for updating a routing table for an adjacent node incorporated in the own node based on the routing information. 3. The router according to claim 1, wherein, when the router is located on both sides of the separated failed node, the built-in routing table of the failed node is replaced with the routing table of the own node. A router comprising means for updating with routing information based on contents and transmitting the routing information to an adjacent node facing the router. 4. Each router constituting a node of the ring-shaped redundant network always holds the routing information of two adjacent nodes in addition to the routing information of its own node. A router node characterized in that nodes located on both sides disconnect the failed node from the network, reconfigure the ring network, and substitute the function of the failed node using the built-in routing information of the failed node. Obstacle avoidance method. 5. The router node failure avoiding method according to claim 4, wherein the routing information of the own node operated by the own node is transmitted to the working transmission line, and the routing information taken from the working transmission line is stored in the standby system. A method of avoiding a failure of a router node, wherein each node updates a routing table of its own node and a routing table of an adjacent node incorporated in its own node by transmitting the transmission table to a transmission line. 6. The method for avoiding a failure of a router node according to claim 4, wherein the routers of the nodes located on both sides of the separated failed node transmit routing information of the built-in failed node. A method for updating a router node based on routing information of its own node and transmitting the routing information to an adjacent node facing the node.