CN101155119B - Method and device for confirming boundary node of autonomous system and its path computing method - Google Patents

Abstract

The invention discloses a method for determining a border node of an autonomous system. A router running an external border gateway protocol in the autonomous system notifies its internal neighbors of external connection information; the internal neighbors establish and maintain external connection sessions according to the external connection information Summary information table; query the external connection session summary information table to obtain the border nodes of the autonomous system. The invention also discloses a corresponding device and a method for calculating traffic engineering paths between autonomous systems. The scheme of the invention adopts a dynamic notification method, solves the problem of determining the border nodes, avoids complicated manual configuration, and facilitates the calculation of traffic engineering paths between autonomous systems.

Description

A method, device and path calculation method for determining autonomous system boundary nodes

technical field

The invention relates to the field of communication technology, in particular to a method, a device and a path calculation method for determining an autonomous system boundary node.

Background technique

With the continuous expansion of the network, the continuous growth of data traffic, and the increasingly complex business, the existing network is overwhelmed. People are increasingly unable to tolerate this Best Effort (Best Effort) transmission mode, and there is an urgent need for network optimization. .

TE (Traffic Engineering, traffic engineering) emerged as the times require. It focuses on the optimization of the overall performance of the network. Its main goal is to provide efficient and reliable network services conveniently, optimize the use of network resources, and optimize network traffic. This is divided into two levels: one is traffic-oriented, that is, it focuses on how to improve the service quality of the network; the other is resource-oriented, that is, it focuses on how to optimize the use of network resources, and the most important thing is the effective use of bandwidth resources. With the expansion of the deployment scope and the development of GMPLS (Generalized Multi-Protocol Label Switching, Generalized Multi-Protocol Label Switching) technology, inter-domain traffic engineering will span different operators and cover GMPLS networks.

CSPF (Constraint-based Shortest Path First) calculation is an important part of MPLS (Multiple Protocol Label Switch) and GMPLS traffic engineering, multi-AS (Autonomous System, autonomous system) Path calculation is very complicated and may require the cooperation of computing entities in different ASs to complete it together. The path computation method based on the PCE (Path Computation Element) model can be applied to inter-domain traffic engineering. When PCC (Path Computation Client, path calculation client), such as ingress LSR (ingress Label Switching router, ingress label switching router), when establishing an LSP path, it sends a request to PCE, and the request information contains destinations and various constraints, etc. Basic information, PCE calculates the path that satisfies the constraints in the request according to the topology synchronized with the network, TED (Traffic Engineering Database, traffic engineering database) and other information, and returns to PCC through the response message as the ERO ( Explicit Route Object, explicit path object) parameter, the calculation result can contain precise nodes (a router) and loose nodes (a network segment, an area, an autonomous system). The calculation range that a PCE is responsible for is generally an autonomous system, because the scope of the diffusion TED information of IGP (Internal Gateway Protocol) is an autonomous system; when the destination of the calculation request is another autonomous system, a different PCEs between autonomous systems cooperate to complete the calculation of a path. The PCE is not limited to a specific implementation form, and can be implemented in a router or a designated server.

Path calculation within an autonomous system can have one or more PCEs. When there is only one PCE, it is a centralized calculation method. All calculation requests within the autonomous system are sent to this PCE. When there are multiple PCEs, All computing requests inside the autonomous system can be distributed to different PCEs to realize computing load sharing, so as to reduce the possibility of failure of computing requests due to PCE blocking. Before the PCC sends a calculation request, it needs to know the computing capabilities of each PCE, such as which QOS (Quality of Service, service quality) computing capabilities it has, whether it can calculate the path for protecting links and computing load sharing, TED synchronization capabilities and Speed, etc., so as to select a PCE to send the calculation request.

In the PCE model, two basic protocols are required: the PCE discovery protocol, which is used by the PCC to discover the existence and computing capabilities of the PCE. When the PCC has a computing request, it selects the appropriate PCE to send the computing request for computing according to the information obtained by the protocol; PCECP (PCE Communication Protocol, PCE Communication Protocol), is mainly responsible for sending calculation requests and response information between PCC and PCE. When PCC sends calculation requests, it contains various constraints. When a PCE calculation requires assistance from other PCEs, PCE The communication protocol support is also required between them.

The PCE model is also applicable to traffic engineering between Areas and autonomous systems. When calculating the TE path between areas, the PCE nodes in different areas cooperate to complete the calculation of an end-to-end TE path. The PCE in each area is only responsible for the calculation of the path in the area, and the next path will be calculated by the PCE node that has completed the calculation. The task is sent to the PCE in the downstream area through the PCECP protocol for calculation. After the calculation is completed by the PCE in the area where the destination of the path is located, the result is returned to the upstream PCE through the PCECP protocol. The upstream PCE will complete the splicing of the path and repeat the previous process until it is returned to the PCE in the area where the source address of the path is located. , the PCE obtains a complete inter-area TE path, and returns the calculation result to the requesting PCC through the PCECP protocol.

For the path calculation between autonomous systems in the PCE environment, the BRPC (Reverse Fallback Path Computation) algorithm is used in the prior art. Assume that the ASs passed through from head (head node) to tail (tail node) are: AS(1), AS(2)...AS(N), wherein tail is located in AS(N). Each AS has a boundary router, which is divided into BN-en (entry Boundary Node, entry boundary node), BN-ex (exitBoundary Node, exit boundary node), where BN-en is the boundary connected to the previous AS on the path Router, BN-ex is a border router connected to the last AS on the path. The PCE of AS(N) calculates the shortest path from all its BN-en to tail, and returns the result to the PCE of AS(N-1); the PCE of AS(N-1) uses this result, AS(N-1 ) and the link TE information between BN-ex of AS(N-1) and BN-en of AS(N), can calculate the shortest path from all BN-en to tail, and then return the result to The PCE of AS(N-2); the shortest path from all BN-en to tail of AS(N-2) is calculated by the PCE of AS(N-2)...so it goes back to AS(2 ) PCE returns its calculation result to the PCE of AS(1), then the PCE of AS(1) can calculate the shortest path from head to tail.

The above-mentioned algorithm can calculate the optimal path and the equivalent path, but it does not solve the problem of how to determine BN-en and BN-ex in the PCE of each AS; unit) to obtain the connectivity information with the external AS through static manual configuration. When the border node changes, it needs to be manually modified; in addition, after static configuration, when there are multiple BN-ex in the same AS, the Calculate the shortest path through all BN-ex, even if there is a link failure between one or some of the BN-ex and other AS, the path calculation will still be performed through the BN-ex, which can only be found during the calculation process Path not available, waste of time.

Contents of the invention

The present invention adopts the way of dynamic notification to solve the problem of determining the border nodes, avoids complex manual configuration, and facilitates the calculation of flow engineering paths between autonomous systems.

Based on the above purpose, the present invention provides a method for determining the border nodes of an autonomous system, a router running an external border gateway protocol in an autonomous system notifies external connection information to its internal neighbors;

The internal neighbor establishes and maintains an external connection session summary information table according to the external connection information;

Querying the external connection session summary information table to obtain the border nodes of the autonomous system.

Wherein, the external connection information is carried in the border gateway protocol message, and the external connection information includes the ID of the external router corresponding to the external connection and the autonomous system number where the external router is located.

The external connection session summary information table includes the number of the autonomous system where the external router is located, the ID of the external router, and the ID of the router running the external border gateway protocol.

The querying the external connection session summary information table to obtain the border node of the autonomous system includes:

Obtaining the boundary node between the autonomous system and the autonomous system where the external router is located by using the number of the autonomous system where the external router is located as an index;

Or, using the number of the autonomous system where the external router is located and the ID of the external router as an index, acquire the border nodes connected to the external routers in the autonomous system.

The router running the external border gateway protocol in the autonomous system advertises external connection information to its internal neighbors, including: when the router running the external border gateway protocol establishes an internal connection with the internal neighbor, it notifies the internal neighbor of a new Adding the external connection information of the router running the external border gateway protocol; and/or, when the router running the external border gateway protocol establishes an external connection, it notifies its internal neighbors of adding the external connection information; and/or , when an external connection of the router running the external BGP fails, notify its internal neighbors to delete the external connection information.

At the same time, the present invention also provides a method for calculating traffic engineering paths between autonomous systems, where the path establishment request located in the first autonomous system initiates a path establishment request to the requested party in the Nth (N>=2) autonomous system;

After receiving the request, the path calculation unit of the autonomous system where the requested party resides queries its external connection session summary information table to determine the boundary nodes that have external connections between the Nth autonomous system and the N-1th autonomous system, and calculates the boundary a path from the node to the requested party;

The path calculation unit sends the calculation result to the path calculation unit of the N-1th autonomous system, and the path calculation unit of the N-1th autonomous system queries its external connection session summary information table to determine the A border node having an external connection with the N-2th autonomous system, calculating a path from the border node to the requested party;

Calculation is rolled back to the first autonomous system in turn, and the path between the requester and the requested party is calculated.

Wherein, the external connection session summary information table of each path computation unit includes an external router ID having an external connection with the autonomous system where the external router is located, an autonomous system where the external router is located, and a local router ID having an external connection with the external router.

In addition, the present invention also provides a border gateway protocol BGP node device, including:

BGP message processing unit: used to receive BGP connection messages sent by other BGP nodes, and transmit them to the BGP connection management unit for processing; and used to carry the EBGP Session Summary NLRI generated by the EBGP Session Summary NLRI processing unit in the BGP message, advertising external connection information to internal neighbors of the BGP node device;

BGP connection management unit: used to receive the BGP connection message sent by the BGP message processing unit, return the response message, and establish and maintain the BGP neighbor table according to the BGP connection message;

EBGP Session Summary NLRI processing unit: used to generate EBGP Session Summary NLRI according to the BGP neighbor table;

Storage unit: storing the BGP neighbor table.

In yet another embodiment, the present invention also provides a BGP node device, including:

BGP message processing unit: used to receive BGP messages sent by other BGP node devices, and transmit the BGP connection message to the BGP connection management unit; and transmit the BGP message carrying the EBGP SessionSummary NLRI information to the EBGP Session Summary NLRI processing unit;

BGP connection management unit: used to receive the BGP connection message sent by the BGP message processing unit, and return the response message; and establish and maintain the BGP neighbor table according to the BGP connection message;

EBGP Session Summary NLRI processing unit: used to receive the BGP message transmitted by the BGP message processing unit, extract the EBGP Session Summary NLRI, and establish and maintain the EBGP Session Summary Table according to the EBGP Session Summary NLRI;

Storage unit: used to store the BGP neighbor table and the EBGP Session SummaryTable.

By adopting the solution of the present invention, the boundary nodes of the autonomous system are determined through the method of dynamic notification, which greatly improves the processing efficiency compared with the method of static manual configuration of the PCE; moreover, the change of the boundary nodes can be updated in time by using the dynamic notification, If the border node link fails or the node itself fails, the PCE can avoid using the node for path calculation during the path calculation process, thereby saving processing time.

Description of drawings

Figure 1 is a PCE model diagram between autonomous systems;

Fig. 2 is a BGP node device figure in a specific embodiment of the present invention;

Fig. 3 is a BGP node device diagram in another specific embodiment of the present invention;

Fig. 4 is a BGP node apparatus diagram in yet another specific embodiment of the present invention;

FIG. 5 is a simplified schematic diagram of an EBGP session summary table in an embodiment of the present invention;

6 is a simplified schematic diagram of the EBGP Session Summary Table of PCE1 in the embodiment of the present invention;

7 is a simplified schematic diagram of the EBGP Session Summary Table of PCE2 in the embodiment of the present invention;

Fig. 8 is a simplified schematic diagram of the EBGP Session Summary Table of PCE 3 in the embodiment of the present invention.

Detailed ways

The core idea of the present invention is to realize the determination of AS boundary nodes by means of dynamic notification.

In the specific implementation process of the present invention, the inter-ASPCE (inter-AS PCE can also calculate the path in the autonomous system where it is located) that is responsible for calculating the path between autonomous systems needs to know all the EBGP (External Border Gateway Protocol, External Border Gateway Protocol) of the AS where it is located. BN-en and BN-ex can be clearly known only if the external Border Gateway Protocol (BGP) connection information is provided. That is to say, Inter-AS PCE needs to establish IBGP (Internal Border Gateway Protocol, internal border gateway protocol) connection with all ASBR (Autonomous System Border Router, autonomous system border router), and ASBR will link the connection information between this AS and external AS, That is, the EBGP connection information is notified to the inter-AS PCE in the local AS. In this process, the BGP protocol needs to be extended.

RFC 2858 defines MP-BGP (Multi Protocol Border Gateway Protocol, Multi-Protocol Border Gateway Protocol), which enables BGP to advertise other network layer information in addition to IPv4 routes, such as defining Multiprotocol Reachable NLRI (Multiprotocol Reachable NLRI) Network Layer Reachability Information) and Multiprotocol Unreachable NLRI (Multiprotocol Unreachable Network Layer Reachability Information) are two BGP attributes to carry the network layer information of these layers.

Among them, the Multiprotocol Reachable NLRI format is shown in Table 1,

Table 1

The format of Multiprotocol Unreachable NLRI is shown in Table 2.

Table 2

Embodiments of the present invention define a new NLRI--EBGP Session Summary NLRI (EBGP Session Summary Network Layer Reachability Information) on the basis of the message formats shown in Table 1 and Table 2.

Among them, the EBGP Session Summary NLRI format used to announce the new EBGP connection information is based on Table 1, and the Network Layer Reachability Information (variable) field is extended to the format shown in Table 3; at the same time, it is also necessary to extend the AFI (Address FamilyIdentifier , Address Family Identifier), SAFI (Subsequent Address Family Identifier, Subsequent Address Family Identifier) to support EBGP Session Summary NLRI, that is, to define new AFI and SAFI numbers to distinguish from IANA (Internet Assigned NumberAuthority, Internet Address Assignment Organization) has defined number. Among them, the Length field indicates the length of the EBGP Session Summary NLRI; the Router ID field indicates the ID of the external neighbor of the ASBR, and the AS number field indicates the number of the AS where the external neighbor is located. The extended message can indicate which external neighbors an ASBR in the local autonomous system has established EBGP connections with.

table 3

The format of the EBGP Session Summary NLRI message used to announce the deletion of EBGP connection information is based on the format shown in Table 2, extending the Withdrawn Routs (variable) field to the format shown in Table 3; where the Length field represents the EBGP Session Summary The length of the NLRI; the Router ID field indicates the external neighbor ID that lost the EBGP connection with the ASBR, and the AS number field indicates the AS number where the external neighbor is located. The extended message can indicate which external neighbor an ASBR in the local autonomous system has lost the EBGP connection with.

Inter-AS PCE establishes and maintains EBGP Session Summary Table (EBGP Session Summary Table) locally according to EBGP Session Summary NLRI information. As shown in Figure 5, the table is a linked list, indexed by the external AS number that is connected to the AS where the Inter-AS PCE is located. Each external AS number corresponds to several Routers, indicating that the external AS has an EBGP connection with the local The router ER (External Router, external router); each of the above-mentioned ERs corresponds to several Routers, indicating the local ASBR (that is, the AS where the Inter-AS PCE is located) that has an EBGP connection with the border router of the external AS.

Of course, AS numbers 100, 200...600 and Router numbers 1, 2, 3, and 4 in Fig. 5 are just examples, and other numbers can also be used, so the numbers used in Fig. 5 are not used to limit the scope of the present invention .

According to the EBGP Session Summary Table, with the AS number as the index, the Inter-AS PCE can find out the local routers that have EBGP connections with the external AS, so as to easily find the ASBRs that are connected to the external AS in the local AS, as shown in Figure 5 , in the AS where the Inter-AS PCE is located, Router1 has an EBGP connection with ER1 in AS100, and Router2 and Router3 have EBGP connections with ER2 in AS100.

The specific operations for establishing and maintaining the EBGP Session Summary Table are as follows:

(1) During the process of establishing an IBGP (Internal Border Gateway Protocol) connection between ASBR and a BGP speaker, if ASBR finds that the BGP speaker has MP-BGP capability and supports extended AFI and SAFI, then after the IBGP connection is established , the ASBR sends an EBGP SessionSummary NLRI message notifying the newly added EBGP connection information to the BGP speaker, and the message carries the EBGP connection information of the ASBR;

After receiving the message, the BGP speaker adds the EBGP connection information carried in it to the EBGP Session Summary Table;

(2) When the ASBR establishes an EBGP connection with an external AS, the EBGP connection information is carried in the EBGP Session Summary NLRI message that notifies the new EBGP connection information, and notifies all its MP-BGP capabilities and support The IBGP neighbors of the extended AFI and SAFI, that is, the BGP speaker that has an IBGP connection relationship with the ASBR;

After receiving the message, the IBGP neighbor adds the EBGP connection information carried in it to the EBGPSession Summary Table;

(3) When an EBGP connection of the ASBR fails, send an EBGPSession Summary NLRI message notifying to delete the EBGP connection information to all its IBGP neighbors that have MP-BGP capabilities and support extended AFI and SAFI;

After receiving the message, the IBGP neighbor deletes the corresponding entry in the EBGP Session Summary Table according to the EBGP connection information carried in it, and completes the update of the EBGP Session Summary Table.

In the above process, ASBR can use its BGP neighbor table to judge which of its IBGP neighbors have MP-BGP capability,

Further detailed description will be given below in conjunction with the accompanying drawings and specific embodiments.

As shown in Figure 1, there are three ASs, namely AS100, AS200, and AS300. Each of these three ASs has an inter-AS PCE, which is correspondingly PCE1, PCE2, and PCE3. The neighbors of R2 are PCE1, R1, and R4. PCE1 and R1 are in AS100, and R4 is in AS200. Suppose the BGP peer Address of R1 is 1.1.1.1, which does not support extended AFI and SAFI; the BGP peer Address of PCE1 is 100.1.1.1, which supports extended AFI and SAFI; the BGP peer Address of R4 is 4.4.4.4, which does not support extended AFI , SAFI.

Then, the BGP neighbor table of R2 can be in the form of Table 4:

BGP peer Address AS No. Peer MP-BGP Capability 1.1.1.1 AS100 <Ipv4,unicast> 100.1.1.1 AS100 <Ipv4, unicast><Ipv4, EBGP Session SummaryReceive Capability> 4.4.4.4 AS200 <Ipv4,unicast>

Table 4

Certainly, Table 4 is only used for illustration, and cannot be used to limit the scope of the present invention, and other information may also be included in the BGP neighbor table. The BGP neighbor table can be dynamically updated, and the dynamic update belongs to the prior art, and will not be repeated here.

The following example illustrates the establishment and maintenance process of the EBGP Session Summary Table on PCE1:

Assuming that R2 (ASBR) has established an EBGP connection with AS200 when establishing an IBGP connection with PCE1, after R2 establishes an IBGP connection with PCE3, it is found that PCE1 has MP-BGP capability and supports extended AFI, SAFI, then R2 sends an EBGP Session Summary NLRI message to PCE1 to announce the new EBGP connection information, which carries the EBGP connection information of R2; after receiving the message, PCE1 adds the EBGP connection information carried in it to In EBGP Session Summary Table.

Assuming that R2 establishes an IBGP connection with PCE1 before establishing an EBGP connection with AS200, the EBGP connection information is carried in the EBGP Session SummaryNLRI message notifying the new EBGP connection information, and notified to PCE1; after receiving the message, PCE1, Add the EBGP connection information carried in it to the EBGP Session Summary Table.

Assuming that the EBGP connection between R2 and AS200 fails, it sends an EBGP Session Summary NLRI message to PCE1 to notify PCE1 of deleting the EBGP connection information; after receiving the message, PCE1 deletes the corresponding table in the EBGP Session Summary Table according to the EBGP connection information carried in it item to complete the update of the EBGP Session Summary Table.

According to the connection relationship shown in Figure 1, the EBGP Session Summary Table of PCE1 is shown in Figure 6.

The EBGP Session Summary Table of PCE2 is shown in Figure 7.

The EBGP Session Summary Table of PCE3 is shown in Figure 8.

Assume that R1 (BGP speaker) of AS100 needs to establish a TE LSP to R9. The specific implementation process is as follows:

Step1, R1 sends a request to establish a TE LSP to PCE1;

Step2, PCE1 searches the BGP routing table and sends the request to PCE2 of AS200;

Step3, PCE2 sends the request to PCE3;

Step4, PCE3 serves AS300 where R9 is located. Therefore, after receiving the above request, it starts path calculation. The specific calculation process is as follows:

(1) According to the EBGP Session Summary Table of PCE3 shown in Figure 8, PCE3 determines that the BN-en of AS300 to which AS200 belongs is R7 and R8, then PCE3 calculates the paths from R7 and R8 to R9 respectively and calculates the The result is returned to PCE2;

(2) After PCE2 receives the calculation result sent by PCE3, according to the EBGPSession Summary Table of PCE2 shown in Figure 7, it is determined that the BN-en of AS200 to which AS100 belongs is R4 and R5; for R7 and R8 , BN-ex is R6, then calculate the paths from R4 and R5 to R6 respectively, and then calculate R4, R5 to R9 according to the calculation results sent by PCE3, the TE link information between R6-R7, and R6-R8 path, and return the calculation result to PCE1;

(3) After PCE1 receives the calculation result sent by PCE2, according to the EBGPSession Summary Table of PCE1 shown in Figure 6, it determines that the BN-ex of the AS100 it belongs to for R4 and R5 is R2 and R3 respectively, and calculates R1 to R2 respectively , the path from R1 to R3, and then combine the calculation results sent by PCE2 and the TE link information between R2-R4 and R3-R5 to calculate the shortest path from R1 to R9.

In the above path calculation process, because PCE1 is an inter-AS PCE serving AS100, it only needs to determine the egress border nodes leading to AS200, that is, R2 and R3, instead of searching for the ingress border nodes; similarly, PCE3 It is only necessary to determine the ingress boundary node leading to AS200.

Correspondingly, in a specific embodiment of the present invention, a BGP node device for sending EBGP connection information is shown in Figure 2, including:

BGP message processing unit: used to receive BGP connection messages sent by other BGP nodes, and transmit them to the BGP connection management unit for processing; and used to carry the EBGP Session Summary NLRI generated by the EBGP Session Summary NLRI processing unit in the BGP message, According to the BGP neighbor table, advertise to BGP neighbors with MP-BGP capability and supporting extended <AFI, SAFI>;

BGP connection management unit: used to receive the BGP connection message sent by the BGP message processing unit, return the response message, and establish and maintain the BGP neighbor table according to the BGP connection message;

EBGP Session Summary NLRI processing unit: used to generate EBGP Session Summary NLRI according to the BGP neighbor table;

Storage unit: storing the BGP neighbor table.

In another embodiment, a BGP device receiving EBGP connection information is shown in Figure 3, which has MP-BGP capability and supports extended AFI and SAFI, including:

BGP message processing unit: used to receive BGP messages sent by other BGP node devices, and transmit the BGP connection message to the BGP connection management unit; and transmit the BGP message carrying the EBGP SessionSummary NLRI information to the EBGP Session Summary NLRI processing unit;

BGP connection management unit: used to receive the BGP connection message sent by the BGP message processing unit, and return the response message; and establish and maintain the BGP neighbor table according to the BGP connection message;

EBGP Session Summary NLRI processing unit: used to receive the BGP message transmitted by the BGP message processing unit, extract the EBGP Session Summary NLRI, and establish and maintain the EBGP Session Summary Table according to the EBGP Session Summary NLRI;

Storage unit: used to store the BGP neighbor table and the EBGP Session SummaryTable.

Figure 4 shows another embodiment of the present invention, which is a BGP node device capable of sending EBGP connection information and receiving EBGP connection information, including:

BGP packet processing unit: used to receive BGP packets sent by other BGP node devices, and transmit BGP connection packets to the BGP connection management unit; transmit BGP packets carrying EBGP Session Summary NLRI information to EBGP Session Summary NLRI for processing unit; it can also be used to carry the EBGP Session SummaryNLRI generated by the EBGP Session Summary NLRI processing unit in the BGP message, and notify the BGP neighbors with MP-BGP capability and support extended AFI and SAFI according to the BGP neighbor table;

BGP connection management unit: used to receive the BGP connection message sent by the BGP message processing unit, and return the response message; and establish and maintain the BGP neighbor table according to the BGP connection message;

EBGP Session Summary NLRI processing unit: used to receive the BGP message sent by the BGP message processing unit, extract the EBGP Session Summary NLRI, establish and maintain the EBGP Session Summary Table according to the EBGP Session Summary NLRI; correspondingly, it can also be used according to the BGP neighbor The table generates EBGP Session Summary NLRI;

Storage unit: used to store the BGP neighbor table and the EBGP Session SummaryTable.

It can be seen from the above specific embodiments that using the method of dynamic notification to determine the boundary nodes of the autonomous system greatly improves the processing efficiency compared with the method of static manual configuration of the PCE; moreover, the use of dynamic notification can update the boundary in time For node changes, if the border node link fails or the node itself fails, PCE can avoid using this node for path calculation during the path calculation process, thereby greatly saving processing time.

The above is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any person familiar with the technology can easily think of changes or replacements within the technical scope disclosed in the present invention. , should be covered within the protection scope of the present invention.

Claims (10)

1. a method of calculating flux engineering route between autonomous system is characterized in that, the request that the requesting party sets up the path to the Requested Party initiation of N autonomous system is set up in the path that is positioned at first autonomous system, and said N is greater than 2;

After the path-calculating element of said Requested Party place autonomous system receives described request; Inquire about its outside connection session summary info table and confirm that N autonomous system and N-1 autonomous system have the outside boundary node that is connected, and calculate the path of said boundary node to said Requested Party;

Said path-calculating element sends to the aforementioned calculation result path-calculating element of N-1 autonomous system; Its outside connection session summary info table of path-calculating element inquiry of said N-1 autonomous system confirms with the N-2 autonomous system the outside boundary node that is connected is arranged in the N-1 autonomous system, calculates the path of said boundary node to said Requested Party;

Rollback calculates first autonomous system successively, calculates the path between requesting party and the Requested Party.

2. the method for claim 1 is characterized in that, said outside connection session summary info table is an External BGP EBGP session abstract, sets up and safeguards that said EBGP session abstract comprises:

The router of the operation External BGP in the autonomous system is to the outside link information of its inner neighbor advertisement;

Said inner neighbours set up according to said outside link information and safeguard outside connection session summary info table.

3. method as claimed in claim 2; It is characterized in that; Said EBGP session abstract is a chained list, is numbered index with outside autonomous system, the outside router that has EBGP to be connected with this autonomous system in this outside autonomous system of the corresponding expression of each outside autonomous system numbering; The local router that the corresponding expression of each said outside router and said outside router have EBGP to be connected.

4. like claim 2 or 3 described methods, it is characterized in that the router of the operation External BGP in the said autonomous system comprises to the outside link information of its inner neighbor advertisement:

When the router of said operation External BGP is set up inner the connection with said inner neighbours, increase the outside link information of the router of said operation External BGP newly to said inner neighbor advertisement;

Or

When the router of said operation External BGP is set up outside a connection, to the newly-increased said outside link information of its inner neighbor advertisement.

5. method as claimed in claim 4; It is characterized in that; Network layer reachability information field to multi-protocols accessibility Network layer reachability information is expanded, and address family identifier and subsequent address family identifier are expanded, and uses the newly-increased said outside link information of message announcement after expanding; Wherein, the numbering of the Address-Family Identifier of the message after said expansion symbol and subsequent address family identifier is different from the numbering of Internet Assigned Number Authority definition.

6. like claim 2 or 3 described methods, it is characterized in that the router of the operation External BGP in the said autonomous system comprises to the outside link information of its inner neighbor advertisement:

An outside connection when losing efficacy of the router of said operation External BGP, delete said outside link information to its inner neighbor advertisement.

7. method as claimed in claim 6 is characterized in that, the route field of cancelling of multi-protocols inaccessibility Network layer reachability information is expanded, and uses the said outside link information of message announcement deletion after expanding.

8. like claim 1,2 or 3 described methods; It is characterized in that the outside connection session summary info table of said each path-calculating element comprises and the autonomous system at the outside outside router ID that is connected of its place autonomous system existence, said outside router place, the local router ID that is connected with said outside router existence outside.

9. a Border Gateway Protocol (BGP) node apparatus is characterized in that, comprising:

BGP message process unit: be used to receive the BGP connection message that other BGP nodes send, send BGP connection management unit to and handle; And be used for the EBGP Session Summary NLRI that External BGP session summary Network layer reachability information EBGP Session Summary NLRI processing unit generates is carried on the BGP message, to the outside link information of the inside of said BGP node apparatus neighbor advertisement;

BGP connection management unit: be used to receive the BGP connection message that the BGP message process unit transmits, the echo reply message, and set up according to said BGP connection message and to safeguard the bgp neighbor table;

EBGP Session Summary NLRI processing unit: be used for generating EBGP Session Summary NLRI according to the bgp neighbor table;

Memory cell: store said bgp neighbor table.

10. a Border Gateway Protocol (BGP) node apparatus is characterized in that, comprising:

BGP message process unit: be used to receive the BGP message that other BGP node apparatus are sent, send the BGP connection message to BGP connection management unit; And the BGP message that will wherein carry EBGP Session Summary NLRI information sends EBGP Session Summary NLRI processing unit to;

BGP connection management unit: be used to receive the BGP connection message that the BGP message process unit transmits, echo reply message; And set up according to said BGP connection message and to safeguard the bgp neighbor table;

EBGP Session Summary NLRI processing unit: be used to receive the BGP message that the BGP message process unit transmits; Extract EBGP Session Summary NLRI, set up according to said EBGP Session Summary NLRI and safeguard External BGP session abstract EBGP Session Summary Table;

Memory cell: be used to store said bgp neighbor table and said EBGP Session Summary Table.

Images