WO2014183628A1 - 一种多域路径计算失败的处理方法及路径计算单元pce - Google Patents

一种多域路径计算失败的处理方法及路径计算单元pce Download PDF

Info

Publication number
WO2014183628A1
WO2014183628A1 PCT/CN2014/077313 CN2014077313W WO2014183628A1 WO 2014183628 A1 WO2014183628 A1 WO 2014183628A1 CN 2014077313 W CN2014077313 W CN 2014077313W WO 2014183628 A1 WO2014183628 A1 WO 2014183628A1
Authority
WO
WIPO (PCT)
Prior art keywords
domain
pce
domain sequence
path calculation
path
Prior art date
Application number
PCT/CN2014/077313
Other languages
English (en)
French (fr)
Inventor
卢刚
曹徐平
Original Assignee
中兴通讯股份有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 中兴通讯股份有限公司 filed Critical 中兴通讯股份有限公司
Priority to ES14797496T priority Critical patent/ES2753215T3/es
Priority to EP14797496.8A priority patent/EP3026850B1/en
Priority to US14/907,558 priority patent/US20160164715A1/en
Publication of WO2014183628A1 publication Critical patent/WO2014183628A1/zh

Links

Classifications

    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L12/00Data switching networks
    • H04L12/64Hybrid switching systems
    • H04L12/6418Hybrid transport
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L41/00Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
    • H04L41/06Management of faults, events, alarms or notifications
    • H04L41/0654Management of faults, events, alarms or notifications using network fault recovery
    • H04L41/0659Management of faults, events, alarms or notifications using network fault recovery by isolating or reconfiguring faulty entities
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L41/00Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
    • H04L41/12Discovery or management of network topologies
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L45/00Routing or path finding of packets in data switching networks
    • H04L45/12Shortest path evaluation

Definitions

  • the present invention relates to a path computation technique, and provides a processing method and system for multi-domain path computation failure.
  • BACKGROUND In a multi-layer multi-domain network using the GMPLS protocol, each node can be divided into different domains according to its switching capability and rate (also can be said to be divided into different layers).
  • AS autonomous system
  • each domain In a multi-domain network partitioned by AS (autonomous system), each domain is equipped with a corresponding PCE (Path Computation Unit), and a path spanning multiple domains is calculated by PCE.
  • PCE Power Computation Unit
  • the entry-to-departure relationship between the domain and the domain and the same domain is established through the node, and the node-to-node path calculated according to the PCE is called a link.
  • RFC5441 proposes a PCE-based BRPC (Reverse Recursive Path Calculation Method): As shown in Figure 1, after the source node and the destination node are clarified, the first domain PCE will be selected according to the topology. A rough path consisting of domains (ie, domain sequence AS1-AS2-AS3), and then the path computation request PCReq to the PCE of AS2 and AS3 is sent according to the sequence of the domain sequence. Then, after receiving the PCReq from the tail domain PCE, it starts to calculate specifically all possible paths between the destination node and the tail domain boundary entry node (ie, the virtual shortest path tree VSPT), and returns the direction according to the domain sequence return direction.
  • the first domain PCE will be selected according to the topology.
  • a rough path consisting of domains ie, domain sequence AS1-AS2-AS3
  • the path computation request PCReq to the PCE of AS2 and AS3 is sent according to the sequence of the domain sequence.
  • the tail domain PCE
  • PCRep path calculation response
  • PCE2 receiving PCRep starts to calculate and generate VSPT of AS2, and attaches the VSPT of AS2 to PCReq and forwards it to upstream PCE1 until the source node
  • PCE1 of the domain confirms the VSPT tree of AS1, it finally determines all the specific paths from the source node to the destination node, and finally selects an optimal path from these specific paths.
  • the tail domain PCE3 fails during the calculation; 2) the intermediate domain PCE2 is Failure during the calculation; 3) The first domain PCE1 failed during the calculation; 4) A failure occurred between the adjacent two domains.
  • RFC5440 shows that when the path calculation fails, the NO-PATH object can be carried in the failure response of PCRep to indicate that the calculation fails.
  • the IRO object records the node or link or domain that must be included in the path, and the XRO object records the node that is excluded.
  • METRIC object records information such as the measurement criteria to determine the best path to the destination, but does not propose the actual PCE processing method for BRPC calculation failure, such as the strategy adopted by the first domain PCE after determining the path calculation failure.
  • an embodiment of the present invention provides a method for processing a multi-domain path calculation failure, which is used in a first-domain path calculation unit PCE, and includes the following steps: acquiring a problem link that causes a path calculation failure of the first domain sequence Deleting the problem link from the first topology to obtain a second topology; determining a second domain sequence based on the second topology; initiating path computation for the second domain sequence.
  • the problem link is a problem link of the downstream PCE or a problem link of the first domain PCE; wherein the problem chain of the downstream PCE is transmitted by the downstream PCE; the problem link of the first domain PCE is The step of determining the second domain sequence according to the second topology includes: determining, according to the second topology, a candidate candidate domain sequence from the source node to the destination node; determining the Where the problem link is located; when the problem link occurs in an intermediate domain, the domain sequence containing the intermediate domain is excluded from the candidate domain sequence, and then the shortest candidate is selected among the remaining candidate domain sequences The domain sequence is used as the second domain sequence; if the candidate domain sequence is a domain sequence including the intermediate domain, then the shortest domain sequence is directly selected as the second domain sequence in the candidate domain sequence; when the first domain or the tail domain or If the problem link occurs between two adjacent domains, a shortest domain sequence is selected as the second domain sequence in the candidate domain sequence.
  • the method further includes: when the first domain PCE fails to calculate a path according to the first domain sequence, transmitting a notification message to the downstream PCE according to the first domain sequence, so that the downstream PCE that receives the notification message is received. Release the resources used to perform path calculations.
  • the step of initiating the path calculation for the second domain sequence includes: sending a path calculation request for the second domain sequence to the downstream PCE according to the second domain sequence; wherein the path calculation request includes the problem The link, such that when the downstream PCE performs path computation for the second domain sequence, the problem link can be excluded according to the path computation request.
  • the embodiment of the present invention further provides a processing method for multi-domain path calculation failure, which is used in the intermediate domain path calculation unit PCE, and includes the following steps: When the intermediate domain PCE fails to calculate the path according to the first domain sequence, the calculation fails. The problem link is passed to the first domain PCE in accordance with the first domain sequence. When the intermediate domain PCE calculates the path failure according to the first domain sequence, the method further includes: transmitting a notification message to the downstream PCE according to the first domain sequence, so that the downstream of the notification message is received.
  • the PCE releases the resources used for path computation.
  • the present invention further includes: receiving a path calculation request for the second domain sequence initiated by the first domain PCE; wherein the path calculation request includes the problem link; when performing the path calculation of the second domain sequence, The problem link in the path calculation request is excluded.
  • the embodiment of the present invention further provides a processing method for multi-domain path calculation failure, which is used in the trailing domain path calculation unit PCE, and includes the following steps: When the trailing domain PCE calculates a path failure according to the first domain sequence, the calculation fails. The problem link is passed to the upstream PCE in accordance with the first domain sequence.
  • the method further includes: receiving a path calculation request for the second domain sequence initiated by the first domain PCE; wherein the path calculation request includes the problem link; when performing the path calculation of the second domain sequence, The problem link in the path calculation request is excluded.
  • the embodiment of the present invention further provides a first domain path calculation unit PCE, including: The problem link determining unit is configured to determine all problem links that cause the path calculation failure of the first domain sequence to occur; the deleting unit is configured to delete all the problem links from the first topology to obtain the second topology; the domain sequence determining unit And being configured to determine a second domain sequence according to the second topology; and an initiating unit to initiate path computation for the second domain sequence.
  • the problem link includes: a problem link of the downstream PCE or a problem link of the first domain PCE; wherein the problem chain of the downstream PCE is transmitted by the downstream PCE; the problem link of the first domain PCE
  • the path sequence determining unit is configured to perform the path calculation by using the first domain PCE.
  • the domain sequence determining unit includes: a first determining subunit, configured to determine, according to the second topology, a candidate domain sequence that is feasible from the source node to the destination node.
  • a second determining subunit configured to determine a location where the problem link is located; a first selecting subunit, configured to exclude the inclusion of the problem link in the candidate domain sequence when the problematic link occurs in an intermediate domain a domain sequence of the intermediate domain, and then selecting a shortest domain sequence as the second domain sequence among the remaining candidate domain sequences; if the candidate domain sequence is a domain sequence including the intermediate domain, directly in the candidate domain sequence Selecting a shortest domain sequence as the second domain sequence; the second selecting subunit is configured to select the problem link between the first domain or the tail domain or the adjacent two domains, and select one of the candidate domain sequences The shortest domain sequence is used as the second domain sequence.
  • the first domain PCE further includes: a first notification unit, configured to: when the first domain PCE fails to calculate a path according to the first domain sequence, send a notification message to the downstream PCE according to the first domain sequence, so that the notification message is received
  • the downstream PCE of the notification message releases the resources used for path computation.
  • the initiating unit is specifically configured to: Transmitting a path calculation request for the second domain sequence to the downstream PCE according to the second domain sequence; wherein the path calculation request includes the problem link, so that the downstream PCE performs a path for the second domain sequence When calculating, the problem link can be excluded according to the path calculation request.
  • the embodiment of the present invention further provides an intermediate domain path calculation unit PCE, including: a first delivery unit, configured to: when the path is calculated according to the first domain sequence fails, the problem link that causes the calculation to fail according to the first The domain sequence is passed to the first domain PCE.
  • the intermediate domain PCE further includes: a notification unit, configured to: when the intermediate domain PCE fails to calculate the path according to the first domain sequence, deliver a notification message to the downstream PCE according to the first domain sequence, so that the notification is received The downstream PCE of the message releases the resources used for path computation.
  • the intermediate domain PCE further includes: a first receiving unit, configured to receive a path calculation request for the second domain sequence initiated by the first domain PCE; wherein the path calculation request includes the problem link; The unit, when performing the path calculation of the second domain sequence, excludes the problem link in the path calculation request.
  • An embodiment of the present invention further provides a trailing domain path calculating unit PCE, including: a second transmitting unit, configured to cause a problem that the calculation fails when the trailing domain PCE fails to calculate a path according to the first domain sequence The link is delivered to the upstream PCE in accordance with the first domain sequence.
  • the tail domain PCE further includes: a second receiving unit, configured to receive a path calculation request for the second domain sequence initiated by the first domain PCE; wherein the path calculation request includes the problem link;
  • the unit when performing the path calculation of the second domain sequence, excludes the problem link in the path calculation request.
  • FIG. 1 is a schematic diagram of a reverse recursive path calculation method BRPC in the prior art
  • FIG. 2 is a schematic diagram of steps of a processing method for multi-domain path calculation failure in a first domain PCE according to the present invention
  • FIG. 4 is a schematic diagram showing the steps of a method for processing a multi-domain path calculation failure for a tail domain PCE in the present invention
  • FIG. 5 is a schematic diagram of a method for processing a multi-domain path calculation failure for a tail domain PCE
  • FIG. 10 is a schematic structural diagram of a first domain PCE in the present invention
  • FIG. 11 is a schematic structural diagram of an intermediate domain PCE according to the present invention
  • a processing method for multi-domain path calculation failure includes a plurality of path calculation units PCE, and the plurality of PCEs can calculate a path of the source node to the destination node based on the reverse recursive path calculation method BRPC.
  • the method includes the following steps: Step 21: Obtain a problem link that causes a path calculation failure of the first domain sequence; Step 22: Delete the problem link from the first topology to obtain a second topology; Step 23, according to the second topology Determining a second domain sequence; Step 24, initiating a path computation for the second domain sequence.
  • the above method re-initiates the calculation of the second domain sequence after the PCE fails to calculate the path of the first domain sequence.
  • the problem link is a problem link of the downstream PCE or a problem link of the first domain PCE; wherein the problem chain of the downstream PCE is transmitted by the downstream PCE;
  • the problem link of the first domain PCE is obtained by path calculation by the first domain PCE; in the current path computation field, a corresponding PCE is configured for each domain.
  • the process of successfully calculating the multi-domain path is roughly as follows: The tail domain PCE first starts the path calculation. When the tail domain PCE successfully calculates the path that it is responsible for, it sends a path calculation response to the upstream adjacent PCE.
  • the neighboring PCE After receiving the path calculation response, the neighboring PCE starts to calculate the path it is responsible for until the first domain PCE successfully completes the path calculation.
  • the path calculation response is sent to the upstream PCE to inform the calculation result, so that the upstream PCE discards the path calculation task to be performed. Therefore, the problem link in this embodiment is derived from the first domain PCE. , either from a PCE downstream.
  • the problem link for the downstream PCE can be provided by the downstream PCE; the problem link for the first domain PCE is directly calculated by its own path.
  • the step 23 specifically includes: Step 231, determining, according to the new topology, a candidate candidate domain sequence from the source node to the destination node; Step 232, determining the link that includes the problem a problem domain; Step 233, when the problem link occurs in an intermediate domain, the domain sequence including the intermediate domain is excluded from the candidate domain sequence, and then the shortest candidate is selected among the remaining candidate domain sequences The domain sequence is used as the second domain sequence; if the candidate domain sequence is a domain sequence including the intermediate domain, then the shortest domain sequence is directly selected as the second domain sequence in the candidate domain sequence; Step 234, when the first domain or If the problem link occurs between the tail domain or the adjacent two domains, a shortest domain sequence is selected as the second domain sequence in the candidate domain sequence.
  • the source node can reach the destination node through multiple candidate domain sequences.
  • the domain sequence capable of avoiding the intermediate domain is preferentially selected as the second domain sequence to improve the probability of success of the next path calculation; if the intermediate domain cannot be avoided, the candidate domain sequence is directly selected.
  • the shortest domain sequence is used as the second domain sequence to ensure the efficiency of the next path computation.
  • the tail field where the first domain or the destination node is located as the source node cannot be replaced in the domain sequence, so the shortest domain sequence is selected directly in the candidate domain sequence as the second domain sequence.
  • the problem link appears between two adjacent domains, considering that there may be other available inter-domain links between the two adjacent domains, it is also directly in the candidate domain sequence after avoiding the problem link.
  • the shortest domain sequence is selected as the second domain sequence.
  • the existing PCE does not immediately release the resources occupied by these calculations after the path calculation is successful. It can be seen from the above that if the path calculation of the first domain PCE fails, the downstream PCE has no meaning for the path from the source node to the destination node even if the path calculation is successfully completed, so as to avoid wasting resources for the downstream PCE.
  • the method further includes the step 25: when the first domain PCE calculates a path failure according to the first domain sequence, transmitting a notification message to the downstream PCE according to the first domain sequence, so that the notification is received.
  • the downstream PCE of the notification message releases the resources used for path computation.
  • the second domain sequence can successfully avoid the problem link appearing in the first domain sequence, in the path calculation specifically for the second domain sequence, the downstream PCE may waste resources again in the domain. The link is calculated.
  • the step 24 specifically includes: sending a path calculation request for the second domain sequence to the downstream PCE according to the second domain sequence; wherein the path calculation request The problem link is included to ensure that the downstream PCE can exclude the problem link according to the path calculation request when performing path calculation for the second domain sequence.
  • the path calculation request is included to ensure that the downstream PCE can exclude the problem link according to the path calculation request when performing path calculation for the second domain sequence.
  • the embodiment of the present invention further provides a processing method for the multi-domain path calculation failure of the intermediate domain PCE, including: Step 31: When the intermediate domain PCE calculates a path failure according to the first domain sequence The problem link that causes the calculation to fail is transmitted to the first domain PCE according to the first domain sequence; the foregoing method may enable the first domain PCE to delete the problem link of the intermediate domain from the first topology to obtain the second topology, and A second domain sequence is determined according to the second topology, and a path computation for the second domain sequence is ultimately initiated. In addition, due to the support of some of the subsequent functions, the existing PCE will not immediately release the resources occupied by these calculations after the path calculation is successful.
  • the method may further include: Step 32: Send a notification message to the downstream PCE according to the first domain sequence, so that the downstream PCE that receives the notification message is released.
  • the method for performing the path calculation in addition, in order to avoid repeatedly calculating the problem link when performing the path calculation of the second domain sequence, as shown in FIG.
  • the method may further include: Step 33: Receive a path calculation request for the second domain sequence initiated by the first domain PCE, where the path calculation request includes the problem link, and step 34, when performing path calculation of the second domain sequence, exclude The problem link in the path calculation request is dropped.
  • an embodiment of the present invention further provides a processing method for a multi-domain path calculation failure of a tail domain PCE, including: Step 41: When a tail domain PCE calculates a path failure according to the first domain sequence The problem link that causes the calculation to fail is delivered to the upstream PCE in accordance with the first domain sequence.
  • the foregoing method may enable the first domain PCE to delete the problem link of the tail domain from the first topology, obtain a second topology, and determine a second domain sequence according to the second topology, and finally initiate path calculation for the second domain sequence.
  • the method may further include: Step 42: Receive path calculation for the second domain sequence initiated by the first domain PCE The path calculation request includes the problem link; Step 43: When performing the path calculation of the second domain sequence, the problem link in the path calculation request is excluded.
  • the problem occurs in the intermediate domain.
  • the second domain sequence can successfully avoid the intermediate domain.
  • six domains are mainly given (first domain AS1, intermediate domain AS2, intermediate domain AS2). , intermediate domain AS3', intermediate domain AS4' and tail domain AS3), the corresponding PCE of each domain is first domain PCE1, intermediate domain PCE2, intermediate domain PCE3, intermediate domain PCE3, intermediate domain PCE4', and tail domain PCE4 (not shown) Draw in the middle).
  • SN is the source node
  • DN is the destination node
  • BN1..BN10 and ⁇ .. BN3' are boundary nodes.
  • the SN can reach the DN through the domain sequence AS1-AS2-AS3, and its specific links are: BN1-BN2-BN3-BN4-DN and ⁇ - ⁇ 2'- ⁇ 3'-BN4-DN; or SN through the domain sequence AS1-AS2'-AS2-AS3 arrives at the DN.
  • the specific link is: BN5-BN6-BN7-BN8-BN9-BN10-DN; or through AS1-AS2-AS3' -AS4'-AS3 arrives at the DN. It is assumed that the links between all the adjacent two domains (that is, the inter-domain links shown by the solid lines in the figure) are in a normal state. After the steps of calculating the inter-domain links are omitted, the main processes are as follows:
  • the SN sends a path calculation request PCReq to PCE1 to calculate a cross-domain path from SN to DN;
  • the PCE1 After receiving the path calculation request PCReq, the PCE1 calculates the shortest domain sequence AS1-AS2-AS3 according to its first topology, and sets the IRO object in the PCReq to AS1, AS2, AS3, and then forwards it to PCE2;
  • PCE2 continues to forward the PCReq request to PCE3;
  • PCE3 as the tail domain PCE starts to perform BRPC calculation and successfully obtains the virtual shortest path tree VSPT3 (ie BN4-DN);
  • PCE3 initiates the path calculation corresponding PCRep, and returns VSPT3 to PCRep and then returns to PCE2; 506.
  • PCE2 starts to perform BRPC calculation after receiving PCRep, since BN2-BN3 and ⁇ 2'- ⁇ 3' are unavailable, the path calculation fails;
  • PCE2 sends a notification message PCNtf to PCE3, notifying PCE3 to release the resources occupied by computing VSPT3;
  • PCE2 puts the problem link into the IRO of the PCRep (ie the link of all BN2-BN3 and the link of ⁇ 2'- ⁇ 3') and sets the NO-PATH object in the PCRep to indicate that the calculation failed; finally PCRep Passed to PCE1;
  • PCE1 knows that the downstream PCE calculation fails according to the NO-PATH object in the PCRep, abandons the path calculation to be performed, and obtains the problem link of BN2-BN3 and ⁇ 2'- ⁇ 3' from the IRO object in the PCRep, and then ⁇ 2 - ⁇ 3 and ⁇ 2'- ⁇ 3' are deleted from their first topology, resulting in a second topology; 510.
  • PCE1 determines candidate domain sequences AS1-AS2'-AS2-AS3 and AS1- based on its second topology
  • AS2-AS3'-AS4'-AS3 it is found that AS2 cannot be excluded from the domain sequence, so a shortest domain sequence AS 1 -AS2'-AS2-AS3 is selected from the candidate domain sequence as the second domain sequence, and Path calculation for AS 1 -AS2'- AS2-AS3;
  • PCE2 forwards the PCReq to PCE3; 514.
  • PCE3 starts the path calculation as the tail domain and gets the new VSPT3 (BN10-DN);
  • PCE3 forwards the new VSPT3 to PCRep and forwards it to PCE2;
  • PCE2 performs path calculation after receiving PCRep, and successfully obtains VSPT2 (BN8-BN9);
  • PCE2 also enters VSPT2 into PCRep and transmits PCRep carrying VSPT2 and VSPT3 to PCE2';
  • PCE2' After receiving the PCRep, PCE2' performs path calculation and successfully obtains VSPT2' (BN6-BN7);
  • PCE2' enters VSPT2' into VSPT3 and returns PCRep carrying VSPT2', VSPT2 and VSPT3 to PCE1;
  • PCE1 completes the path calculation of the first domain after receiving the PCRep, obtains VSPT1 (SN-BN5), and finally determines the path of SN to lj DN through VSPT1, VSPT2, VSPT2', and VSPT3;
  • PCE1 replies the PCRep (the path carrying the SN to the DN) to the SN node.
  • the second domain sequence can successfully avoid the intermediate domain.
  • five domains are given (first domain AS1, intermediate domain AS2, intermediate domain AS2', intermediate domain AS3', and tail domain AS3.
  • the corresponding PCEs of each domain are PCE1, PCE2, PCE3, PCE2', PCE3' (not shown in the figure).
  • SN is the source node
  • DN is the destination node
  • BN1..BN7 is the boundary node. It is assumed that the links between all the adjacent two domains (that is, the inter-domain links shown by the solid lines in the figure) are in a normal state. After the steps of calculating the inter-domain links are omitted, the main processes are as follows:
  • the SN sends a PCReq to the PCE1 to calculate a cross-domain path from the SN to the DN;
  • the PCE1 After receiving the PCReq, the PCE1 calculates the first domain sequence as AS1-AS2-AS3 according to the first topology, and sets the IRO in the PCReq to AS 1, AS2, AS3. And forward the PCReq request to PCE2;
  • PCE2 continues to forward the PCReq request to PCE3;
  • PCE3 as the tail domain PCE after receiving the PCReq, the first to perform the path calculation, and successfully get VSPT3 (BN7-DN);
  • PCE3 passes VSPT3 into PCRep and passes it to PCE2; 606.
  • PCE2 performs path calculation after receiving PCRep, and path calculation fails because BN2-BN3 is unavailable;
  • PCE2 sends PCNtf to PCE3, notifying PCE3 to release the resources occupied by computing VSPT3;
  • PCE2 enters the link of BN2-BN3 into the IRO object of PCRep, and sets the NO-PATH object in PCRep to indicate the calculation failure; finally passes PCRep to PCE1;
  • PCE1 After receiving the PCRep, PCE1 knows the path calculation to be performed after the downstream PCE calculation fails according to the NO-PATH object, and obtains the BN2-BN3 problem link from the IRO object in the PCRep, and then BN2-BN3 Deleting from its first topology, obtaining a second topology;
  • PCE1 determines a unique candidate domain sequence AS1-AS2'-AS3 according to its second topology, uses AS1-AS2'-AS3 as the second domain sequence, and initiates path computation for AS1-AS2'-AS3;
  • PCE2' forwards the PCReq request to PCE3;
  • PCE3 performs path calculation after receiving PCReq, and successfully obtains new VSPT3; 613.
  • PCE3 initiates PCRep and forwards PCRep carrying VSPT3 to PCE2'; 614.
  • AS2' performs path calculation after receiving PCRep, successfully obtains VSPT2 ' ;
  • PCE2' enters VSPT2' into PCRep and returns PCRep carrying VSPT2' and VSPT3 to PCE1;
  • PCE1 completes the path calculation of the first domain after receiving PCRep, obtains VSPT1, and finally determines the path of SN to DN through VSPT1, VSPT2', and VSPT3; 617.
  • PCE1 records the path of SN to DN into PCRep and returns to SN node.
  • PCE2 forwards the PCReq request to PCE3; 704.
  • PCE3 acts as the tail domain PCE performs the path calculation first after receiving the PCReq, and the calculation fails because BN4-DN is not available;
  • PCE3 initiates a PCRep, carries the BN4-DN link in the IRO object of the PCRep, and sets the NO-PATH object of the PCRep to indicate that the calculation failed;
  • PCE2 forwards PCRep to PCE1; 707.
  • PCE1 knows that the downstream PCE calculation fails according to the NO-PATH object, discards the path calculation to be performed, and obtains the BN4-DN problem chain from the IRO object in PCRep. After the BN4-DN is deleted from its first topology, the second topology is obtained;
  • PCE1 uses the shorter AS1-AS2'-AS3 as the second domain sequence and initiates the AS1 -AS2'-AS3 path calculation;
  • PCE2' forwards the PCReq request to PCE3;
  • the tail domain PCE3 After receiving the PCReq, the tail domain PCE3 performs the new path calculation first; wherein, according to the XRO information in the PCRep, the BN4-DN is determined to be the problem link, and after the exclusion, the BN7-DN is directly calculated;
  • PCE3 enters VSPT3 (BN7-DN) into PCRep and passes it to PCE2';
  • PCE2' performs path calculation after receiving PCRep, and successfully obtains VSPT2' (BN5-BN6);
  • PCE2' forwards VSPT2' to PCRep and forwards it to PCE1;
  • PCE1 completes the BRPC calculation of the first domain, obtains lj VSPT1, and finally obtains the path of SN to DN through VSPTK VSPT2 and VSPT3; 716.
  • PCE1 replies the PCRep response (the path carrying the SN to the DN) to the SN node.
  • the SN sends a PCReq request to PCE1 to calculate a cross-domain path from SN to DN;
  • PCE2 continues to forward the PCReq request to PCE3;
  • PCE3 as the tail domain PCE, performs the path calculation first after receiving the PCReq, and obtains VSPT3 (BN4-DN);
  • PCE3 generates PCRep, and records VSPT3 into PCRep and sends it to PCE2; 806. After receiving PCRep initiated by PCE3, PCE2 performs path calculation to obtain VSPT2 (BN2-BN3);
  • PCE2 enters VSPT2 into PCRep and forwards the PCRep carrying VSPT2 and VSPT3 to PCE1;
  • PCE1 After receiving the PCRep, PCE1 starts the path calculation. Because the link of SN-BN1 is unavailable, the calculation fails. 809. PCE1 sends a PCNtf message to PCE2, informing PCE2 to release the resources occupied by VSPT2.
  • PCE2 releases the resources used to calculate VSPT2
  • PCE2 forwards the PCNtf message to the PCE3, and notifies the PCE3 to release the resource occupied by the VSPT3. 812.
  • the PCE3 releases the resource occupied by the VSPT3. 813.
  • PCE1 deletes SN-BN1 from its first topology to obtain a second topology; and determines, according to the second topology, that the second domain sequence is AS1-AS2'-AS3, and then initiates a path for AS1-AS2'-AS3. Calculation
  • PCE2' forwards PCReq request to PCE3 ;
  • PCE3 performs the path calculation again after receiving the PCReq, and still obtains VSPT3 as BN4-DN;
  • PCE3 generates PCReP, and VSPT3 is entered into PCReP, and then forwards PCRep carrying VSPT3 to PCE2';
  • PCE2' performs path calculation after receiving PCRep, and obtains VSPT2' (BN5-BN6); 819. PCE2' enters VSPT2' into PCReP, and then returns PCRep carrying VSPT2' and VSPT3 to PCE1;
  • PCE1 performs path calculation after receiving PCRep, and obtains VSPT1 (SN-AS1);
  • PCE1 obtains the path of the SN-DN according to VSPT1, VSPT2' and VSPT3.
  • BN1-BN2, BN3-BN4, BN7-BN5, and BN6-BN4 are inter-domain links (that is, links between domains and domains).
  • inter-domain links such as BN1-BN2— It is generally calculated by the PCE close to the upstream (ie PCE1), and PCE2 is responsible for calculating the inter-domain link BN3-BN4.
  • the specific path calculation process is as follows:
  • the SN sends a PCReq request to PCE1 to calculate a cross-domain path from SN to DN;
  • PCE2 continues to forward the PCReq request to PCE3; 904.
  • the tail domain PCE3 performs the path computation first after receiving the PCReq, and obtains VSPT3; 905.
  • PCE3 replies to the PCRep of VSPT3 to PCE2;
  • PCE2 acts as the intermediate domain PCE, and performs path calculation to obtain VSPT2. However, in the calculation, it is found that the inter-domain link BN3-BN4 is unavailable, so the calculation fails.
  • PCE2 sends a PCRep message to PCE1;
  • PCE1 After receiving the PCRep, PCE1 knows the path calculation to be performed after the downstream PCE calculation fails according to the NO-PATH object, and obtains the problem link BN3-BN4 according to the IRO object of PCRep, and then BN3-BN4 from its own The first topology is deleted, and the second topology is obtained;
  • PCE1 determines a unique candidate domain sequence AS1-AS2'-AS3 according to its second topology, uses AS1-AS2'-AS3 as the second domain sequence, and initiates path calculation for AS1-AS2'-AS2-AS3;
  • PCE3 begins to perform path calculation after receiving PCReq, and still obtains BN4-DN;
  • PCE3 enters VSPT3 into PCRep, and then sends PCRep carrying VSPT3 to PCE2';
  • PCE2' performs path calculation after receiving PCRep, and obtains VSPT2';
  • PCE1 performs path calculation after receiving PCRep, and obtains VSPT1;
  • PCE1 finally determines the path of SN-DN from VSPT2' and VSPT3 in PCRep and according to VSPT1 calculated by itself. It should be noted that in the present embodiment, the problem link is not necessarily required to be carried in the PCRep and the PCReq, and the dedicated information may be separately configured and transmitted between the PCEs. In addition, based on the embodiment, an upper limit value may be set to limit the number of times the first domain PCE initiates recalculation. In summary, the above embodiment has the following advantages:
  • an embodiment of the present invention further provides a first domain path calculation unit PCE, including: a problem link determining unit, configured to determine all problem links that cause a path calculation failure of the first domain sequence; , the method is configured to remove all the problem links from the first topology, to obtain a second topology; the domain sequence determining unit is configured to determine the second domain sequence according to the second topology; and the initiating unit initiates the second domain sequence Path calculation.
  • a problem link determining unit configured to determine all problem links that cause a path calculation failure of the first domain sequence
  • the method is configured to remove all the problem links from the first topology, to obtain a second topology
  • the domain sequence determining unit is configured to determine the second domain sequence according to the second topology
  • the initiating unit initiates the second domain sequence Path calculation.
  • the problem link includes: a problem link of a downstream PCE or a problem link of a first domain PCE; wherein the problem chain of the downstream PCE is transmitted by a downstream PCE
  • the problem link of the first domain PCE is obtained by path calculation by the first domain PCE; in the current path determination field, a corresponding PCE is configured for each domain.
  • the process of calculating the multi-domain path is as follows: The tail domain PCE first starts the path calculation. When the tail domain PCE successfully calculates the link that it is responsible for, it sends a path calculation response to the upstream adjacent PCE. After receiving the path calculation response, the adjacent PCE starts to calculate the link it is responsible for until the first domain PCE successfully completes the path calculation. When a PCE fails to calculate the link, the upstream PCE is notified of the calculation result by the path calculation response, so that the upstream PCE discards the path calculation task to be performed. Therefore, the problem link in this embodiment is derived from the first domain. PCE, either from a downstream PCE.
  • the problem link for the downstream PCE can be provided by the downstream PCE; the problem link for the first domain PCE is directly calculated by its own path.
  • the domain sequence determining unit includes: The first determining subunit is configured to determine, according to the second topology, a candidate candidate domain sequence from the source node to the destination node.
  • a second determining subunit configured to determine a location where the problem link is located; a first selecting subunit, configured to exclude the inclusion of the problem link in the candidate domain sequence when the problematic link occurs in an intermediate domain a domain sequence of the intermediate domain, and then selecting a shortest domain sequence as the second domain sequence among the remaining candidate domain sequences; if the candidate domain sequence is a domain sequence including the intermediate domain, directly in the candidate domain sequence Selecting a shortest domain sequence as the second domain sequence; the second selecting subunit is configured to select the problem link between the first domain or the tail domain or the adjacent two domains, and select one of the candidate domain sequences The shortest domain sequence is used as the second domain sequence.
  • the source node can reach the destination node through multiple candidate domain sequences.
  • the domain sequence capable of avoiding the intermediate domain is preferentially selected as the second domain sequence to improve the probability of success of the next path calculation; if the intermediate domain cannot be avoided, the candidate domain sequence is directly selected.
  • the shortest domain sequence is used as the second domain sequence to ensure the efficiency of the next path computation.
  • the tail field where the first domain or the destination node is located as the source node cannot be replaced in the domain sequence, so the shortest domain sequence is selected directly in the candidate domain sequence as the second domain sequence.
  • the second topology is generated after avoiding the problem link.
  • the shortest domain sequence is selected directly as the second domain sequence among the calculated candidate domain sequences.
  • the existing PCE does not immediately release the resources occupied by these calculations after the path calculation is successful. It can be seen from the above that if the path calculation of the first domain PCE fails, the downstream PCE does not have any meaning for the path from the source node to the destination node even if the path calculation of the source node to the destination node is successful.
  • the first domain PCE further includes: a first notification unit, configured to: when the first domain chain PCE fails to calculate a path according to the first domain sequence, downstream according to the first domain sequence
  • the PCE delivers a notification message that causes the downstream PCE that received the notification message to release resources for path computation.
  • the second domain sequence can successfully avoid the problem link appearing in the first domain sequence, in the path calculation specifically for the second domain sequence, the PCE may waste resources to calculate the problem link again.
  • the initiating unit is specifically configured to: Transmitting a path calculation request for the second domain sequence to the downstream PCE according to the second domain sequence; wherein the path calculation request includes the problem link, so that the downstream PCE performs a path for the second domain sequence
  • the problem link can be excluded according to the path calculation request.
  • an embodiment of the present invention further provides an intermediate domain path calculation unit PCE, including: a first delivery unit configured to cause the calculation to fail when the path is failed according to the first domain sequence. The problem link is passed to the first domain PCE in accordance with the first domain sequence.
  • the foregoing method may enable the first domain PCE to delete the problem link of the intermediate domain from the first topology, obtain a second topology, and determine a second domain sequence according to the second topology, and finally initiate path calculation for the second domain sequence. .
  • the existing PCE will not immediately release the resources occupied by these calculations after the path calculation is successful. It can be seen from the above that if the path calculation fails in the intermediate domain PCE, the downstream PCE has no meaning for the entire source node to the destination node even if the path calculation is successfully completed, so as to avoid wasting resources for the downstream PCE. As shown in FIG.
  • the intermediate PCE further includes: a notification unit, configured to: when the intermediate domain PCE fails to calculate a path according to the first domain sequence, send a notification message to the downstream PCE according to the first domain sequence, so as to receive The downstream PCE of the notification message releases the resources used for path computation.
  • the intermediate domain PCE may further include: a first receiving unit, configured to receive the target initiated by the first domain PCE a path calculation request of the second domain sequence; wherein the path calculation request includes the problem link; the first exclusion unit, in performing the path calculation of the second domain sequence, excluding the path calculation request Problem link.
  • a notification unit configured to: when the intermediate domain PCE fails to calculate a path according to the first domain sequence, send a notification message to the downstream PCE according to the first domain sequence, so as to receive The downstream PCE of the notification message releases the resources used for path computation.
  • the intermediate domain PCE may further include: a first receiving unit, configured to receive the target initiated by the first domain PCE a path calculation request of the second domain sequence; where
  • an embodiment of the present invention further provides a trailing domain path calculating unit PCE, including: a second transmitting unit, configured to cause the trailing domain PCE to fail when calculating a path according to the first domain sequence
  • the problem link for the calculation failure is passed to the upstream PCE in accordance with the first domain sequence.
  • the foregoing method may enable the first domain PCE to delete the problem link of the tail domain from the first topology, obtain a second topology, and determine a second domain sequence according to the second topology, and finally initiate path calculation for the second domain sequence. .
  • a trailing domain path calculating unit PCE including: a second transmitting unit, configured to cause the trailing domain PCE to fail when calculating a path according to the first domain sequence
  • the problem link for the calculation failure is passed to the upstream PCE in accordance with the first domain sequence.
  • the foregoing method may enable the first domain PCE to delete the problem link of the tail domain from the first topology, obtain a second topology, and determine a second domain
  • the tail domain PCE may further include: a second receiving unit, configured to receive the target initiated by the first domain PCE a path calculation request of the second domain sequence; wherein the path calculation request includes the problem link; and the second exclusion unit, in performing the path calculation of the second domain sequence, excluding the path calculation request Problem link.
  • the first domain PCE, the intermediate domain PCE, and the tail domain PCE are the device embodiments corresponding to the processing method for the multi-domain path calculation failure in the present invention, and the technical effect that the multi-domain path calculation failure processing method can achieve, The first domain PCE, the intermediate domain PCE, and the tail domain PCE in this embodiment can also be achieved.

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Data Exchanges In Wide-Area Networks (AREA)

Abstract

本发明公开一种多域路径计算失败的处理方法及路径计算单元PCE;其中,用于首域PCE的方法包括:获取导致第一域序列发生路径计算失败的问题链路;将问题链路从第一拓扑中删除,得到第二拓扑;根据所述第二拓扑确定第二域序列;发起针对所述第二域序列的路径计算。本发明的方案在PCE按照第一域序列确定源节点到目的节点的路径时,如果有链路导致计算失败,则将该导致计算失败的链路从首域PCE的拓扑中删除,得到第二拓扑,之后根据第二拓扑制定第二域序列,并发起针对第二域序列的路径计算,从而最终确定出一条源节点通向目的节点的路径。

Description

一种多域路径计算失败的处理方法及路径计算单元 PCE 技术领域 本发明涉及路径计算技术, 提供一种多域路径计算失败的处理方法及系统。 背景技术 在使用 GMPLS协议的多层多域网络中, 每个节点可以根据其交换能力和速率的 划分为不同域 (也可说为划分为不同层)。 在采用 AS (自治系统)进行划分的多域网络 中, 每个域均配有对应的 PCE (路径计算单元), 通过 PCE来计算跨越多个域的路径。 其中, 域与域之间、 同一域的进入位置到离开位置均是通过节点建立连接的, 而根据 PCE具体计算出的节点到节点的路径被称作链路。 在现有的路径计算领域里, RFC5441提出了一种基于 PCE的 BRPC (反向递归路 径计算法): 如图 1所示, 在明确源节点和目的节点后, 首域 PCE会根据拓扑选出一 条由域组成的大致的路径 (即域序列 AS1-AS2-AS3 ), 之后根据该域序列的顺序发送 路径计算请求 PCReq至 AS2和 AS3的 PCE。 然后, 从尾域 PCE在接收到 PCReq后, 开始具体计算出由目的节点到尾域边界进入节点之间的所有可能存在的路径 (即虚拟 最短路径树 VSPT), 并按域序列返回方向发送带有尾域 VSPT的 PCRep (路径计算响 应)至上游的 PCE2, 收到 PCRep的 PCE2开始计算并生成 AS2的 VSPT, 并将该 AS2 的 VSPT附在 PCReq后转发至向上游的 PCE1 , 直到源节点所在域的 PCE1确认 AS1 的 VSPT树后, 最终确定出源节点到目的节点的所有具体路径, 最终再从这些具体路 径中选出一条最优的路径。 目前, 在采用 BRPC算法的跨域路径计算的实际应用中, 存在失败的情况, 根据 失败的位置, 至少存在四种失败场景: 1 ) 尾域 PCE3在计算过程中失败; 2) 中间域 PCE2在计算过程中失败; 3 )首域 PCE1在计算过程中失败; 4)相邻两域之间发生失 败。 而 RFC5440 中虽然说明了当路径计算失败时, 可在 PCRep 的失败响应中携带 NO-PATH对象表明计算失败, IRO对象记录路径必须包含的节点或链路或域, XRO 对象记录计算排斥的节点、 链路或域, METRIC对象记录确定到达目的地的最佳路径 的计量标准等信息, 但未提出实际的 PCE针对 BRPC计算失败的处理方法, 如首域 PCE在确定路径计算失败后采取何种策略再次进行路径计算以避开失败的拓扑区域的 方法。 发明内容 本发明实施例要解决的技术问题是提供一种多域路径计算失败的处理方法及路径 计算单元 PCE, 能够在多域路径计算失败后, 重新计算出一条新的路径。 为解决上述技术问题, 本发明的实施例提供一种多域路径计算失败的处理方法, 用于首域路径计算单元 PCE, 包括如下步骤: 获取导致第一域序列发生路径计算失败的问题链路; 将问题链路从第一拓扑中删除, 得到第二拓扑; 根据所述第二拓扑确定第二域序列; 发起针对所述第二域序列的路径计算。 其中, 所述问题链路为下游 PCE的问题链路或首域 PCE的问题链路; 其中, 所述下游 PCE的问题链是由下游 PCE传递得到的; 所述首域 PCE的问题 链路是通过首域 PCE进行路径计算得到的; 其中, 根据所述第二拓扑确定第二域序列的步骤包括: 根据所述第二拓扑, 确定出源节点到目的节点可行的候选域序列; 确定所述问题链路所在的位置; 当某一中间域出现所述问题链路, 则在所述候选域序列中排除掉包含该中间域的 域序列, 之后在剩下的候选域序列中选取一个最短的域序列作为第二域序列; 若所述 候选域序列均为包含该中间域的域序列, 则在该候选域序列直接选取一个最短的域序 列作为第二域序列; 当首域或尾域或相邻两域之间出现所述问题链路, 则在所述候选域序列中选取一 最短的域序列作为第二域序列。 其中, 本方法还包括: 当所述首域 PCE根据所述第一域序列计算路径失败时,按照所述第一域序列向下 游 PCE传递一通知消息, 使接收到所述通知消息的下游 PCE释放掉用于进行路径计 算的资源。 其中, 发起针对所述第二域序列的路径计算的步骤包括: 按照所述第二域序列向下游 PCE发送针对该第二域序列的路径计算请求; 其中, 所述路径计算请求包含所述问题链路,从而使得下游 PCE在针对所述第二域序列进行 路径计算时, 能够根据所述路径计算请求排除掉所述问题链路。 本发明的实施例还提供一种多域路径计算失败的处理方法, 用于中间域路径计算 单元 PCE, 包括如下步骤: 当中间域 PCE根据第一域序列计算路径失败时,将导致该计算失败的问题链路按 照所述第一域序列向首域 PCE传递。 其中, 当中间域 PCE根据第一域序列计算路径失败时还包括: 按照所述第一域序列向下游 PCE传递通知消息, 使接收到所述通知消息的下游
PCE释放掉用于进行路径计算的资源。 其中, 本发明还包括: 接收首域 PCE发起的针对第二域序列的路径计算请求; 其中, 所述路径计算请求 包含所述问题链路; 在进行所述第二域序列的路径计算时, 排除掉所述路径计算请求中的问题链路。 本发明的实施例还提供一种多域路径计算失败的处理方法, 用于尾域路径计算单 元 PCE, 包括如下步骤: 当尾域 PCE根据第一域序列计算路径失败时,将导致该计算失败的问题链路按照 所述第一域序列向上游 PCE传递。 其中, 本方法还包括: 接收首域 PCE发起的针对第二域序列的路径计算请求; 其中, 所述路径计算请求 包含所述问题链路; 在进行所述第二域序列的路径计算时, 排除掉所述路径计算请求中的问题链路。 本发明的实施例还提供一种首域路径计算单元 PCE, 包括: 问题链路确定单元, 设置为确定所有导致第一域序列发生路径计算失败的问题链 路; 删除单元, 设置为将所有问题链路从第一拓扑中删除, 得到第二拓扑; 域序列确定单元, 设置为根据所述第二拓扑确定第二域序列; 发起单元, 发起针对所述第二域序列的路径计算。 其中, 所述问题链路包括: 下游 PCE的问题链路或首域 PCE的问题链路; 其中, 所述下游 PCE的问题链是 由下游 PCE传递得到的; 所述首域 PCE的问题链路是通过首域 PCE进行路径计算得 到的; 其中, 所述域序列确定单元包括: 第一确定子单元, 设置为根据所述第二拓扑, 确定出源节点到目的节点可行的候 选域序列。 第二确定子单元, 设置为确定所述问题链路所在的位置; 第一选取子单元, 设置为当某一中间域出现所述问题链路, 则在所述候选域序列 中排除掉包含该中间域的域序列, 之后在剩下的候选域序列中选取一个最短的域序列 作为第二域序列; 若所述候选域序列均为包含该中间域的域序列, 则在该候选域序列 直接选取一个最短的域序列作为第二域序列; 第二选取子单元, 设置为当首域或尾域或相邻两域之间出现所述问题链路, 则在 所述候选域序列中选取一最短的域序列作为第二域序列。 其中, 所述首域 PCE还包括: 第一通知单元, 设置为当首域 PCE根据所述第一域序列计算路径失败时, 按照所 述第一域序列向下游 PCE传递通知消息, 使接收到所述通知消息的下游 PCE释放掉 用于进行路径计算的资源。 其中, 所述发起单元具体设置为: 按照所述第二域序列向下游 PCE发送针对该第二域序列的路径计算请求; 其中, 所述路径计算请求包含所述问题链路,从而使得下游 PCE在针对所述第二域序列进行 路径计算时, 能够根据所述路径计算请求排除掉所述问题链路。 本发明的实施例还提供一种中间域路径计算单元 PCE, 包括: 第一传递单元, 设置为当根据第一域序列计算路径失败时, 将导致该计算失败的 问题链路按照所述第一域序列向首域 PCE传递。 其中, 所述中间域 PCE还包括: 通知单元, 设置为当中间域 PCE根据第一域序列计算路径失败时, 按照所述第一 域序列向下游 PCE传递一通知消息, 使接收到所述通知消息的下游 PCE释放掉用于 进行路径计算的资源。 其中, 所述中间域 PCE还包括: 第一接收单元, 设置为接收首域 PCE发起的针对第二域序列的路径计算请求; 其 中, 所述路径计算请求包含所述问题链路; 第一排除单元, 在进行所述第二域序列的路径计算时, 排除掉所述路径计算请求 中的问题链路。 本发明的实施例还提供一种尾域路径计算单元 PCE, 其特征在于, 包括: 第二传递单元, 设置为当尾域 PCE根据第一域序列计算路径失败时, 将导致该计 算失败的问题链路按照所述第一域序列向上游 PCE传递。 其中, 所述尾域 PCE还包括: 第二接收单元, 设置为接收首域 PCE发起的针对第二域序列的路径计算请求; 其 中, 所述路径计算请求包含所述问题链路; 第二排除单元, 在进行所述第二域序列的路径计算时, 排除掉所述路径计算请求 中的问题链路。 本发明实施例的方案具有以下优点: 本发明实施例的方案在 PCE按照第一域序列确定源节点到目的节点的路径时,如 果有链路导致计算失败, 则将该导致计算失败的链路从首域 PCE的拓扑中删除, 得到 第二拓扑, 之后根据第二拓扑制定第二域序列, 并发起针对第二域序列的路径计算, 从而最终确定出一条源节点通向目的节点的路径。 附图说明 图 1为现有技术中的反向递归路径计算法 BRPC的示意图; 图 2为本发明中用于首域 PCE的多域路径计算失败的处理方法的步骤示意图; 图 3为本发明中用于中间域 PCE的多域路径计算失败的处理方法的步骤示意图; 图 4为本发明中用于尾域 PCE的多域路径计算失败的处理方法的步骤示意图; 图 5-9为本发明中问题链路出现在不同位置的情况下, 多域路径计算的步骤图; 图 10为本发明中首域 PCE的结构示意图; 图 11为本发明中中间域 PCE的结构示意图; 图 12为本发明中尾域 PCE的结构示意图。 具体实施方式 为使本发明要解决的技术问题、 技术方案和优点更加清楚, 下面将结合附图及具 体实施例进行详细描述。 如图 2所示, 一种多域路径计算失败的处理方法, 包括多个路径计算单元 PCE, 所述多个 PCE能够基于反向递归路径计算法 BRPC计算出源节点通向目的节点的路 径, 包括如下步骤: 步骤 21, 获取导致第一域序列发生路径计算失败的问题链路; 步骤 22, 将问题链路从第一拓扑中删除, 得到第二拓扑; 步骤 23, 根据所述第二拓扑确定第二域序列; 步骤 24, 发起针对所述第二域序列的路径计算。 上述方法在 PCE在第一域序列的路径计算发生失败后,重新发起第二域序列的计 算。 由于第二域序列是根据第二拓扑确定的, 因此可以避开问题链路, 保证了重新进 行路径计算的成功率。 具体地, 在本发明的上述实施例中, 所述问题链路为下游 PCE的问题链路或首域 PCE的问题链路; 其中, 所述下游 PCE的问题链是由下游 PCE传递得到的; 所述首 域 PCE的问题链路是通过首域 PCE进行路径计算得到的; 在目前的路径计算领域中, 都会为每个域配置一个对应的 PCE。 其中, 对于多域 路径计算成功的过程大致为: 尾域 PCE首先开始进行路径计算, 当尾域 PCE成功计 算出自己负责的路径后,会发送一个路径计算响应至上游相邻的 PCE,上游相邻的 PCE 在接收到该路径计算响应后再开始计算其所负责的路径,直至首域 PCE成功完成路径 计算。 当有一个 PCE发生路径计算失败时, 会通过路径计算响应向上游 PCE告知计 算结果, 从而使上游 PCE放弃掉将要执行的路径计算任务, 因此本实施例中的问题链 路要么源自首域 PCE, 要么源自下游的某一 PCE。对于下游 PCE的问题链路则可由下 游 PCE负责提供; 对于首域 PCE的问题链路则直接通过自身的路径计算得到。 具体地, 在本发明的上述实施例中, 步骤 23具体包括: 步骤 231, 根据所述新的拓扑, 确定出源节点到目的节点可行的候选域序列; 步骤 232, 确定含有所述问题链路的问题域; 步骤 233, 当某一中间域出现所述问题链路, 则在所述候选域序列中排除掉包含 该中间域的域序列, 之后在剩下的候选域序列中选取一个最短的域序列作为第二域序 列; 若所述候选域序列均为包含该中间域的域序列, 则在该候选域序列直接选取一个 最短的域序列作为第二域序列; 步骤 234, 当首域或尾域或相邻两域之间出现所述问题链路, 则在所述候选域序 列中选取一最短的域序列作为第二域序列。 在一般情况下, 源节点可以通过多种候选域序列到达目的节点。 当问题链路出现 在中间域时, 考虑到该中间域可能自身就存在一些不稳定的因素, 导致其即便出现在 下次的路径计算中, 还是会有很大的几率出现新的问题链路, 因此在候选域序列中, 优先选择能够避开该中间域的域序列作为第二域序列, 以提高下次路径计算的成功几 率; 如果不能避开该中间域, 则直接在候选域序列中选择最短的域序列作为第二域序 列以保证下次路径计算的效率。 而作为源节点所在的首域或者目的节点所在的尾域由 于在域序列中是不能被取代的, 因此直接在候选域序列中选取最短的域序列作为第二 域序列。 此外, 如果问题链路出现在相邻两域之间, 考虑到这两个相邻域间可能还存 在其他可用的域间链路, 因此也是避开该问题链路后, 直接在候选域序列中选取最短 的域序列作为第二域序列。 此外, 由于为后续的一些功能提供支持, 现有的 PCE在路径计算成功后并不会马 上释放掉这些计算所占用的资源。通过上文可知, 如果首域 PCE发生路径计算失败的 话, 则下游 PCE即便成功完成了各自的路径计算, 但对于整个源节点到目的节点的路 径来说也是没有意义的, 为避免下游 PCE浪费资源, 在本发明的上述实施例中, 还包 括步骤 25: 当所述首域 PCE根据所述第一域序列计算路径失败时,按照所述第一域序列向下 游 PCE传递通知消息, 使接收到所述通知消息的下游 PCE释放掉用于进行路径计算 的资源。 此外, 第二域序列虽然可以成功避开掉出现在第一域序列中的问题链路, 但是在 具体针对第二域序的路径计算中,下游 PCE还是可能会浪费资源再次对其域内的问题 链路进行计算, 因此, 在本发明的上述实施例中, 步骤 24具体包括: 按照所述第二域序列向下游 PCE发送针对该第二域序列的路径计算请求; 其中, 所述路径计算请求包含所述问题链路,从而保证下游 PCE在针对所述第二域序列进行 路径计算时, 能够根据所述路径计算请求排除掉所述问题链路。 此外, 如图 3所示, 本发明的实施例还提供一种用于中间域 PCE的多域路径计算 失败的处理方法, 包括: 步骤 31, 当中间域 PCE根据第一域序列计算路径失败时,将导致该计算失败的问 题链路按照所述第一域序列向首域 PCE传递; 上述方法可以使首域 PCE将中间域的问题链路从第一拓扑中删除,得到第二拓扑, 并根据第二拓扑确定出第二域序列, 最终发起针对所述第二域序列的路径计算。 此外, 由于为后续的一些功能提供支持, 现有的 PCE在路径计算成功后是不会马 上释放掉这些计算所占用的资源。通过上文可知, 如果中间域 PCE发生路径计算失败 的话, 则下游 PCE即便成功完成了各自的路径计算, 但对于整个端到端路径来说也是 没有意义的, 为避免下游 PCE浪费资源, 如图 3所示, 在本发明的上述实施例中, 步 骤 31后还可以包括: 步骤 32,按照所述第一域序列向下游 PCE传递通知消息,使接收到所述通知消息 的下游 PCE释放掉用于进行路径计算的资源; 此外, 为避免在执行第二域序列的路径计算时, 重复计算问题链路, 如图 3所示, 本方法还可以包括: 步骤 33, 接收首域 PCE发起的针对第二域序列的路径计算请求; 其中, 所述路径 计算请求包含所述问题链路; 步骤 34, 在进行所述第二域序列的路径计算时, 排除掉所述路径计算请求中的问 题链路。 此外, 如图 4所示, 本发明的实施例还提供一种用于尾域 PCE的多域路径计算失 败的处理方法, 包括: 步骤 41, 当尾域 PCE根据第一域序列计算路径失败时,将导致该计算失败的问题 链路按照所述第一域序列向上游 PCE传递。 上述方法可以使首域 PCE将尾域的问题链路从第一拓扑中删除, 得到第二拓扑, 并根据第二拓扑确定出第二域序列, 最终发起针对所述第二域序列的路径计算。 此外, 为避免在执行第二域序列的路径计算时, 重复计算问题链路, 如图 4所示, 本方法还可以包括: 步骤 42, 接收首域 PCE发起的针对第二域序列的路径计算请求; 其中, 所述路径 计算请求包含所述问题链路; 步骤 43, 在进行所述第二域序列的路径计算时, 排除掉所述路径计算请求中的问 题链路。 下面将首域 PCE、 中间域 PCE以及尾域 PCE的多域路径计算失败的处理方法结 合进实际应用中, 进行详细描述:
<一>中间域出现问题链路 ( 1 ) 第二域序列能够成功避开掉该中间域 如图 5所示, 主要给出了六个域 (首域 AS1、 中间域 AS2、 中间域 AS2' 、 中间 域 AS3' 、 中间域 AS4' 以及尾域 AS3 ), 各域相应的 PCE分别为首域 PCE1 、 中间域 PCE2、 中间域 PCE3、 中间域 PCE3、 中间域 PCE4' 、 尾域 PCE4 (未在图中画出)。 SN为源节点, DN为目的节点, BN1..BN10以及 ΒΝΓ .. BN3'均为边界节点。 理论上 讲, SN可通过域序列 AS1-AS2-AS3到达 DN,其具体链路为: BN1-BN2-BN3-BN4-DN 以及 ΒΝΓ-ΒΝ2'-ΒΝ3' -BN4-DN; 或者 SN通过域序列 AS1-AS2'-AS2-AS3到达 DN, 其具体链路为: BN5-BN6-BN7-BN8-BN9-BN10-DN ; 再或者通过 AS1- AS2-AS3' -AS4'-AS3 到达 DN。 假设所有相邻两域之间的链路 (即图中实线所示的域间链路) 均处于正常状态, 省略掉计算域间链路的步骤后, 其主要过程如下:
501. SN向 PCE1发送路径计算请求 PCReq, 计算 SN到 DN的跨域路径;
502. PCE1收到路径计算请求 PCReq后, 根据自身的第一拓扑计算出最短的域序 列 AS1-AS2-AS3 , 并将 PCReq中的 IRO对象设置为 AS1, AS2, AS3后转发给 PCE2;
503. PCE2继续将 PCReq请求转发给 PCE3;
504. PCE3作为尾域 PCE在接收 PCReq后, 开始执行 BRPC计算并成功得到虚 拟最短路径树 VSPT3 (即 BN4-DN);
505. PCE3发起路径计算相应 PCRep, 并将 VSPT3录入 PCRep后回复给 PCE2; 506. PCE2收到 PCRep后开始执行 BRPC计算,由于 BN2-BN3 以及 ΒΝ2'-ΒΝ3' 均不可用, 路径计算失败;
507. PCE2向 PCE3发送通知消息 PCNtf, 通知 PCE3释放掉计算 VSPT3所占用 的资源;
508. PCE2 将问题链路放入 PCRep 的 IRO 中 (即所有 BN2-BN3 的链路以及 ΒΝ2'-ΒΝ3'的链路),并设置 PCRep中的 NO-PATH对象以表明计算失败;最后将 PCRep 传递至 PCE1 ;
509. PCE1根据 PCRep中的 NO-PATH对象得知下游 PCE计算失败, 放弃将要进 行的路径计算, 并从 PCRep中的 IRO对象得到 BN2-BN3和 ΒΝ2'-ΒΝ3' 的问题链路, 之后将 ΒΝ2-ΒΝ3和 ΒΝ2'-ΒΝ3'从自身的第一拓扑删除, 得到第二拓扑; 510. PCE1 根据其第二拓扑确定出候选域序列 AS1-AS2'-AS2-AS3 和 AS1-
AS2-AS3'-AS4'-AS3 , 发现 AS2不能从域序列中排除掉, 因此从候选域序列中选取一 个最短的域序列 AS 1 -AS2'-AS2-AS3作为第二域序列,并发起针对 AS 1 -AS2'- AS2-AS3 的路径计算;
511. PCE1设置 PCReq中的 IR0= AS1-AS2'- AS2-AS3, 并发送给 PCE2' ;
Figure imgf000012_0001
513. PCE2将 PCReq转发给 PCE3; 514. PCE3作为尾域开始执行路径计算, 并得到新的 VSPT3 (BN10-DN);
515. PCE3将新的 VSPT3录入 PCRep后转发至 PCE2;
516. PCE2在收到 PCRep后执行路径计算, 并且成功得到 VSPT2 (BN8-BN9);
517. PCE2同样将 VSPT2录入 PCRep,并将携带有 VSPT2以及 VSPT3的 PCRep 传递至 PCE2' ;
518. PCE2' 收到 PCRep后, 执行路径计算, 并成功得到 VSPT2' (BN6-BN7);
519. PCE2'将 VSPT2'录入至 VSPT3 , 并将携带有 VSPT2'、 VSPT2以及 VSPT3 的 PCRep返回给 PCE1 ;
520. PCE1在收到 PCRep完成首域的路径计算, 得到 VSPT1 ( SN-BN5 ), 并最 后通过 VSPT1、 VSPT2、 VSPT2'、 VSPT3确定 SN至 lj DN的路径;
521. PCE1将 PCRep (携带 SN到 DN的路径) 回复给 SN节点。
(2) 第二域序列能够成功避开掉该中间域 如图 6中所示, 给出了五个域 (首域 AS1、 中间域 AS2、 中间域 AS2' 、 中间域 AS3'以及尾域 AS3 ), 其各域相应的 PCE分别为 PCE1、 PCE2、 PCE3、 PCE2'、 PCE3' (未在图中画出)。 SN为源节点, DN为目的节点, BN1..BN7为边界节点。 假设所有 相邻两域之间的链路 (即图中实线所示的域间链路) 均处于正常状态, 省略掉计算域 间链路的步骤后, 其主要过程如下:
601. SN向 PCE1发送 PCReq, 计算 SN到 DN的跨域路径;
602. PCE1收到 PCReq后根据第一拓扑计算出第一域序列为 AS1-AS2-AS3 , 将 PCReq中的 IRO设置为 AS 1, AS2, AS3。 并将 PCReq请求转发给 PCE2;
603. PCE2继续将 PCReq请求转发给 PCE3;
604. PCE3作为尾域 PCE在接收 PCReq后,率先执行路径计算,并成功得到 VSPT3 (BN7-DN);
605. PCE3将 VSPT3录入 PCRep后传递给 PCE2; 606. PCE2在收到 PCRep后执行路径计算, 由于 BN2-BN3不可用, 路径计算失 败;
607. PCE2向 PCE3发送 PCNtf, 通知 PCE3释放掉计算 VSPT3所占用的资源;
608. PCE2将 BN2-BN3的链路录入 PCRep的 IRO对象中, 并设置 PCRep中的 NO-PATH对象以表明计算失败; 最后将 PCRep传递至 PCE1 ;
609. PCE1在收到 PCRep后, 根据 NO-PATH对象得知下游 PCE计算失败后, 放 弃将要进行的路径计算, 并从 PCRep中的 IRO对象得到 BN2-BN3的问题链路, 之后 将 BN2-BN3从自身的第一拓扑删除, 得到第二拓扑;
610 . PCE1 根据其第二拓扑确定出唯一候选域序列 AS1-AS2'-AS3, 将 AS1-AS2'-AS3作为第二域序列, 并发起针对 AS1-AS2'- AS3的路径计算;
611. PCE1设置 PCReq请求中 IRO=ASl, AS2', AS3 , 并发送给 PCE2' ;
612. PCE2'将 PCReq请求转发给 PCE3 ;
613. PCE3在收到 PCReq后执行路径计算, 成功得到新的 VSPT3; 613. PCE3发起 PCRep, 并将携带 VSPT3的 PCRep转发给 PCE2' ; 614. AS2'收到 PCRep后执行路径计算, 成功得到 VSPT2' ;
615. PCE2'将 VSPT2'录入至 PCRep, 并将携带有 VSPT2'以及 VSPT3的 PCRep 返回给 PCE1 ;
616. PCE1在收到 PCRep完成首域的路径计算,得到 VSPT1 ,并最后通过 VSPT1、 VSPT2'、 VSPT3确定 SN到 DN的路径; 617. PCE1将 SN到 DN的路径录入 PCRep后回复给 SN节点。
<二>尾域出现问题链路 如图 7所示,给出了四个域(首域 AS1、中间域 AS2、中间域 AS2' 、中间域 AS3' 以及尾域 AS3 ), 各域相应的 PCE分别为 PCE1、 PCE2、 PCE2' 、 PCE3'、 PCE3 (未 在图中画出)。 SN为源节点, DN为目的节点, BN1-BN7为域边界节点。 假设所有相 邻两域之间的链路 (即图中实线所示的域间链路) 均处于正常状态, 省略掉计算域间 链路的步骤后, 其主要过程如下: 701. SN向 PCE1发送 PCReq请求, 计算 SN到 DN的跨域路径;
702. PCE1根据第一拓扑计算出第一域序列 AS1-AS2-AS3 , 并将其设置到 PCReq 的 IRO对象中 (即 IR0=AS1, AS2, AS3 ) ;之后将 PCReq请求转发给 PCE2;
703. PCE2将 PCReq请求转发给 PCE3; 704. PCE3作为尾域 PCE在收到 PCReq后率先执行路径计算, 由于 BN4-DN不 可用, 导致计算失败;
705. PCE3发起 PCRep, 将 BN4-DN的链路携带在 PCRep的 IRO对象中, 并设 置 PCRep的 NO-PATH对象以表明计算失败;
706. PCE2向 PCE1转发 PCRep; 707. PCE1在收到 PCRep后, 根据 NO-PATH对象得知下游 PCE计算失败, 放弃 将要进行的路径计算, 并从 PCRep中的 IRO对象得到 BN4-DN的问题链路, 之后将 BN4-DN从自身的第一拓扑删除, 得到第二拓扑;
708 . PCE1 根据其第二拓扑确定出候选域序列 AS1-AS2'-AS3 以及 AS1-AS2'-AS3'-AS3 后, 将较短 AS1-AS2'-AS3 作为第二域序列, 并发起针对 AS1-AS2'-AS3的路径计算;
709. PCE1设置 PCReq中的 IR0=AS1-AS2'-AS3, 并将问题链路 BN4-DN录入 PCReq的 XRO对象中, 之后将设置好的 PCReq发送给 PCE2';
710. PCE2'将 PCReq请求转发给 PCE3 ;
711. 尾域 PCE3在收到 PCReq后, 率先执行新的路径计算;其中, 根据 PCRep中 的 XRO信息确定 BN4-DN为问题链路, 在将其排除后, 直接计算出 BN7-DN;
712. PCE3将 VSPT3 (BN7-DN) 录入 PCRep后传递给 PCE2';
713. PCE2'在接收到 PCRep后执行路径计算, 并成功得到 VSPT2' (BN5-BN6);
714. PCE2'将 VSPT2'录入 PCRep后转发给 PCE1 ;
715. PCE1完成首域的 BRPC计算,得至 lj VSPT1 ,并通过 VSPTK VSPT2、 VSPT3 最终得到 SN到 DN的路径; 716. PCE1将 PCRep响应 (携带 SN到 DN的路径) 回复给 SN节点。
<三>首域出现问题链路 如图 8中所示, 给出了四个域 (首域 AS1、 中间域 AS2、 中间域 AS2' 以及尾域 AS3 ), 各域相应的 PCE分别为 PCE1、 PCE2、 PCE2' 、 PCE3 (未在图中画出)。 SN 为源节点, DN为目的节点, BN1..BN7为域边界节点。 假设所有相邻两域之间的链路
(即图中实线所示的域间链路) 均处于正常状态, 省略掉计算域间链路的步骤后, 其 主要过程如下:
801. SN向 PCE1发送 PCReq请求, 计算 SN到 DN的跨域路径;
802. PCE1根据目前的第一拓扑计算出第一域序列为 AS1-AS2-AS3 ,并设置 PCReq 的 IRO =ASl, AS2, AS3后转发给 PCE2;
803. PCE2继续将 PCReq请求转发给 PCE3;
804. PCE3 作为尾域 PCE, 在收到 PCReq后率先执行路径计算, 得到 VSPT3 (BN4-DN);
805. PCE3生成 PCRep, 并将 VSPT3录入至 PCRep后, 将其发送给 PCE2; 806. PCE2在收到 PCE3发起的 PCRep后,执行路径计算,得到 VSPT2(BN2-BN3 );
807. PCE2将 VSPT2录入至 PCRep中,并将该携带有 VSPT2以及 VSPT3的 PCRep 转发给 PCE1 ;
808. PCE1在收到 PCRep后, 开始执路径计算, 由于 SN-BN1的链路不可用, 计 算失败; 809. PCE1向 PCE2发送 PCNtf消息, 通知 PCE2释放掉计算 VSPT2所占用的资 源;
810. PCE2释放掉计算 VSPT2所占用的资源;
811. PCE2向 PCE3转发 PCNtf消息, 通知 PCE3释放掉计算 VSPT3所占用的资 源; 812. PCE3释放掉计算 VSPT3所占用的资源; 813. PCE1将 SN-BN1从自身的第一拓扑删除, 得到第二拓扑; 并根据第二拓扑 确定出第二域序列为 AS1-AS2'-AS3, 之后发起针对 AS1-AS2'-AS3的路径计算;
814. PCE1设置 PCReq中的 IR0=AS1-AS2'-AS3, 以及将 SN-BN1录入至 PCReq 的 XRO对象中, 之后将设置好的 PCReq发送给 PCE2'; 815. PCE2'将 PCReq请求转发给 PCE3 ;
816. PCE3在收到 PCReq后再次执行路径计算, 并依然得到 VSPT3为 BN4-DN;
817. PCE3生成 PCReP,并 VSPT3录入至 PCReP中,之后将携带 VSPT3的 PCRep 转发给 PCE2' ;
818. PCE2'在收到 PCRep后执行路径计算, 得到 VSPT2' (BN5-BN6); 819. PCE2'将 VSPT2'录入至 PCReP, 之后将携带 VSPT2'以及 VSPT3的 PCRep 返回给 PCE1 ;
820. PCE1在收到 PCRep后进行路径计算, 得到 VSPT1 ( SN-AS1 );
821. PCE1根据 VSPT1、 VSPT2'以及 VSPT3得到 SN-DN的路径。
<四>相邻两域之间出现问题链路 如图 9中所示, 给出了四个 AS (首域 AS1、 中间域 AS2、 中间域 AS2' 以及尾域
AS3 ), 各域相应的 PCE分别为 PCE1、 PCE2、 PCE2' 、 PCE3 (未在图中画出)。 SN 为源节点, DN为目的节点, BN1..BN7为域边界节点。 其中, BN1-BN2、 BN3-BN4 BN7-BN5、 BN6-BN4均为域间链路(即域与域之间的链路), 在目前的 BRPC计算中, 域间链路如 BN1-BN2—般是由靠近上游的 PCE负责计算 (即 PCE1 ), 而 PCE2负责 计算域间链路 BN3-BN4。 其具体的路径计算过程如下:
901. SN向 PCE1发送 PCReq请求, 计算 SN到 DN的跨域路径;
902. PCE1根据第一拓扑计算出第一域序列为 AS1-AS2-AS3 , 并设置 PCReq的 IRO =ASl, AS2, AS3后转发给 PCE2;
903. PCE2继续将 PCReq请求转发给 PCE3; 904. 尾域 PCE3在收到 PCReq后率先执行路径计算, 得到 VSPT3 ; 905. PCE3将 VSPT3的 PCRep回复给 PCE2;
906. PCE2作为中间域 PCE, 进行路径计算, 得到 VSPT2, 但在计算中还发现域 间链路 BN3-BN4不可用, 因此计算失败;
907. PCE2向 PCE3发送 PCNtf消息, 通知 PCE3释放掉计算 VSPT3所占用的资 源; 同时设置 PCRep的 IRO= BN3-BN4, 并设置 PCRep的 NO-PATH对象表明计算失 败;
908. PCE2向 PCE1发送 PCRep消息;
909. PCE1在收到 PCRep后, 根据 NO-PATH对象得知下游 PCE计算失败后, 放 弃将要进行的路径计算, 根据 PCRep 的 IRO对象得到问题链路 BN3-BN4, 之后将 BN3-BN4从自身的第一拓扑删除, 得到第二拓扑;
910 . PCE1 根据其第二拓扑确定出唯一候选域序列 AS1-AS2'-AS3, 将 AS1-AS2'-AS3作为第二域序列, 并发起针对 AS1-AS2'- AS2-AS3的路径计算;
911. PCE1设置 PCReq中的 IR0=AS1-AS2'-AS3,并将 BN3-BN4录入至 XRO中, 之后将设置好的 PCReq发送给 PCE2'; 912. PCE2'将 PCReq请求转发给 PCE3 ;
913. PCE3在收到 PCReq后开始执行路径计算, 依然得到 BN4-DN;
914. PCE3将 VSPT3录入至 PCRep中,之后将携带 VSPT3的 PCRep发给 PCE2';
915. PCE2'在收到 PCRep后执行路径计算, 得到 VSPT2' ;
916. ?。£2'将¥8?丁2'录入至!^«^中,之后将携带 VSPT2'以及 VSPT3的 PCRep 返回给 PCE1 ;
917. PCE1在收到 PCRep后执行路径计算, 得到 VSPT1 ;
918. PCE1从 PCRep中的 VSPT2'以及 VSPT3 , 并根据自身计算到的 VSPT1最 终确定出 SN-DN的路径。 需要指出的是, 在本实施例中, 问题链路并不是必需要搭载在 PCRep和 PCReq 中, 也可以是单独配置专用信息在各 PCE之间传递。 此外, 在本实施例的基础上, 还 可设置一个上限值, 以限制首域 PCE发起重新计算的路径次数。 综上所述, 上述实施例具有以下优点:
1. 由于 PCRep和 PCReq在现有的 BRPC计算中已经存在, 因此不需要再额外生 成信息用来传递问题链路。
2.当中间域出现问题链路时,优先选择不包含该中间域的域序列作为第二域序列, 从而提高了计算第二域序列的路径的成功几率。 此外, 如图 10所示, 本发明的实施例还提供首域路径计算单元 PCE, 包括: 问题链路确定单元, 设置为确定所有导致第一域序列发生路径计算失败的问题链 路; 删除单元, 设置为将所有问题链路从第一拓扑中删除, 得到第二拓扑; 域序列确定单元, 设置为根据所述第二拓扑确定第二域序列; 发起单元, 发起针对所述第二域序列的路径计算。 上述首域 PCE在第一域序列的路径计算发生失败后,重新发起第二域序列的计算。 由于第二域序列是根据第二拓扑确定的, 因此可以避开错误链路, 保证了重新执行路 径计算的成功率。 具体地, 在本发明的上述实施例中, 所述问题链路包括: 下游 PCE的问题链路或首域 PCE的问题链路; 其中, 所述下游 PCE的问题链是 由下游 PCE传递得到的; 所述首域 PCE的问题链路是通过首域 PCE进行路径计算得 到的; 在目前的路径确定领域中, 都会为每个域配置一个对应的 PCE。 其中, 对于多域 路径计算成功的过程大致为: 尾域 PCE首先开始进行路径计算, 当尾域 PCE成功计 算出自己负责的链路后,会发送一个路径计算响应至上游相邻的 PCE,上游相邻的 PCE 在接收到该路径计算响应后再开始计算其所负责的链路,直至首域 PCE成功完成路径 计算。 当有一个 PCE发生链路计算失败时, 会通过路径计算响应向上游 PCE告知计 算结果, 从而使上游 PCE放弃掉将要执行的路径计算任务, 因此本实施例中的问题链 路要么源自首域 PCE, 要么源自下游的某一 PCE。对于下游 PCE的问题链路则可由下 游 PCE负责提供; 对于首域 PCE的问题链路则直接通过自身的路径计算得到。 具体地, 在本发明的上述实施例中, 所述域序列确定单元包括: 第一确定子单元, 设置为根据所述第二拓扑, 确定出源节点到目的节点可行的候 选域序列。 第二确定子单元, 设置为确定所述问题链路所在的位置; 第一选取子单元, 设置为当某一中间域出现所述问题链路, 则在所述候选域序列 中排除掉包含该中间域的域序列, 之后在剩下的候选域序列中选取一个最短的域序列 作为第二域序列; 若所述候选域序列均为包含该中间域的域序列, 则在该候选域序列 直接选取一个最短的域序列作为第二域序列; 第二选取子单元, 设置为当首域或尾域或相邻两域之间出现所述问题链路, 则在 所述候选域序列中选取一最短的域序列作为第二域序列。 在一般情况下, 源节点可以通过多种候选域序列到达目的节点。 当问题链路出现 在中间域时, 考虑到该中间域可能自身就存在一些不稳定的因素, 导致其即便出现在 下次的路径计算中, 还是会有很大的几率出现新的问题链路, 因此在候选域序列中, 优先选择能够避开该中间域的域序列作为第二域序列, 以提高下次路径计算的成功几 率; 如果不能避开该中间域, 则直接在候选域序列中选择最短的域序列作为第二域序 列以保证下次路径计算的效率。 而作为源节点所在的首域或者目的节点所在的尾域由 于在域序列中是不能被取代的, 因此直接在候选域序列中选取最短的域序列作为第二 域序列。 此外, 如果问题链路出现在相邻两域之间, 考虑到这两个相邻域间还可能有 其他可用的域间链路, 因此也是避开该问题链路后, 生成第二拓扑, 直接在计算出的 候选域序列中选取最短的域序列作为第二域序列。 此外, 由于为后续的一些功能提供支持, 现有的 PCE在路径计算成功后并不会马 上释放掉这些计算所占用的资源。通过上文可知, 如果首域 PCE发生路径计算失败的 话, 则下游 PCE即便成功完成了各自的路径计算, 但对于源节点到目的节点的路径来 说也是没有意义的, 为避免下游 PCE浪费资源, 在本发明的上述实施例中, 所述首域 PCE还包括: 第一通知单元, 设置为当首域链 PCE根据所述第一域序列计算路径失败时, 按照 所述第一域序列向下游 PCE传递一通知消息, 使接收到所述通知消息的下游 PCE释 放掉用于进行路径计算的资源。 此外, 第二域序列虽然可以成功避开掉出现在第一域序列中的问题链路, 但是在 具体针对第二域序的路径计算中, PCE还是可能会浪费资源再次对问题链路进行计算, 因此, 在本发明的上述实施例中, 所述发起单元具体设置为: 按照所述第二域序列向下游 PCE发送针对该第二域序列的路径计算请求; 其中, 所述路径计算请求包含所述问题链路,从而使得下游 PCE在针对所述第二域序列进行 路径计算时, 能够根据所述路径计算请求排除掉所述问题链路。 此外,如图 11所示,本发明的实施例还提供一种中间域路径计算单元 PCE,包括: 第一传递单元, 设置为当根据第一域序列计算路径失败时, 将导致该计算失败的 问题链路按照所述第一域序列向首域 PCE传递。 上述方法可以使首域 PCE将中间域的问题链路从第一拓扑中删除,得到第二拓扑, 并根据第二拓扑确定出第二域序列, 最终发起针对所述第二域序列的路径计算。 此外, 由于为后续的一些功能提供支持, 现有的 PCE在路径计算成功后是不会马 上释放掉这些计算所占用的资源。通过上文可知, 如果中间域 PCE发生路径计算失败 的话, 则下游 PCE即便成功完成了各自的路径计算, 但对于整个源节点到目的节点的 路径来说也是没有意义的, 为避免下游 PCE浪费资源, 如图 11所示, 所述中间 PCE 还包括: 通知单元, 设置为当中间域 PCE根据第一域序列计算路径失败时, 按照所述第一 域序列向下游 PCE传递通知消息, 使接收到所述通知消息的下游 PCE释放掉用于进 行路径计算的资源。 此外,为避免在执行第二域序列的路径计算时,重复计算错误链路,如图 11所示, 所述中间域 PCE还可以包括: 第一接收单元, 设置为接收首域 PCE发起的针对第二域序列的路径计算请求; 其 中, 所述路径计算请求包含所述问题链路; 第一排除单元, 在进行所述第二域序列的路径计算时, 排除掉所述路径计算请求 中的问题链路。 此外, 如图 12所示, 本发明的实施例还提供一种尾域路径计算单元 PCE, 包括: 第二传递单元, 设置为当尾域 PCE根据第一域序列计算路径失败时, 将导致该计 算失败的问题链路按照所述第一域序列向上游 PCE传递。 上述方法可以使首域 PCE将尾域的问题链路从第一拓扑中删除, 得到第二拓扑, 并根据第二拓扑确定出第二域序列, 最终发起针对所述第二域序列的路径计算。 此外,为避免在执行第二域序列的路径计算时,重复计算问题链路,如图 12所示, 所述尾域 PCE还可以包括: 第二接收单元, 设置为接收首域 PCE发起的针对第二域序列的路径计算请求; 其 中, 所述路径计算请求包含所述问题链路; 第二排除单元, 在进行所述第二域序列的路径计算时, 排除掉所述路径计算请求 中的问题链路。 显然, 上述首域 PCE、 中间域 PCE以及尾域 PCE为本发明中一种多域路径计算 失败的处理方法对应的装置实施例, 所述多域路径计算失败的处理方法能达到的技术 效果, 本实施例中的首域 PCE、 中间域 PCE以及尾域 PCE也同样能够达到。 工业实用性 本发明实施例的方案在 PCE按照第一域序列确定源节点到目的节点的路径时,如 果有链路导致计算失败, 则将该导致计算失败的链路从首域 PCE的拓扑中删除, 得到 第二拓扑, 之后根据第二拓扑制定第二域序列, 并发起针对第二域序列的路径计算, 从而最终确定出一条源节点通向目的节点的路径。 以上所述是本发明的优选实施方式, 应当指出, 对于本技术领域的普通技术人员 来说, 在不脱离本发明所述原理的前提下, 还可以作出若干改进和润饰, 这些改进和 润饰也应视为本发明的保护范围。

Claims

权 利 要 求 书
1. 一种多域路径计算失败的处理方法, 用于首域路径计算单元 PCE, 包括如下步 骤:
获取导致第一域序列发生路径计算失败的问题链路;
将问题链路从第一拓扑中删除, 得到第二拓扑;
根据所述第二拓扑确定第二域序列;
发起针对所述第二域序列的路径计算。
2. 根据权利要求 1所述的多域路径计算失败的处理方法, 其中, 所述问题链路为 下游 PCE的问题链路或首域 PCE的问题链路;
其中,所述下游 PCE的问题链路是由下游 PCE传递得到的;所述首域 PCE 的问题链路是通过首域 PCE进行路径计算得到的。
3. 根据权利要求 1所述的多域路径计算失败的处理方法, 其中,
根据所述第二拓扑确定第二域序列的步骤包括:
根据所述第二拓扑, 确定出源节点到目的节点可行的候选域序列; 确定所述问题链路所在的位置;
当某一中间域出现所述问题链路, 则在所述候选域序列中排除掉包含该中 间域的域序列, 之后在剩下的候选域序列中选取一个最短的域序列作为第二域 序列; 若所述候选域序列均为包含该中间域的域序列, 则在该候选域序列直接 选取一个最短的域序列作为第二域序列;
当首域或尾域或相邻两域之间出现所述问题链路, 则在所述候选域序列中 选取一最短的域序列作为第二域序列。
4. 根据权利要求 1所述的多域路径计算失败的处理方法,其中,所述方法还包括:
当所述首域 PCE根据所述第一域序列计算路径失败时,按照所述第一域序 列向下游 PCE传递一通知消息, 使接收到所述通知消息的下游 PCE释放掉用 于进行路径计算的资源。
5. 根据权利要求 1所述的多域路径计算失败的处理方法, 其中, 发起针对所述第 二域序列的路径计算的步骤包括: 按照所述第二域序列向下游 PCE发送针对该第二域序列的路径计算请求; 其中, 所述路径计算请求包含所述问题链路, 从而使得下游 PCE在针对所述第 二域序列进行路径计算时, 能够根据所述路径计算请求排除掉所述问题链路。
6. 一种多域路径计算失败的处理方法, 用于中间域路径计算单元 PCE, 包括如下 步骤:
当中间域 PCE根据第一域序列计算路径失败时,将导致该计算失败的问题 链路按照所述第一域序列向首域 PCE传递。
7. 根据权利要求 6所述的多域路径计算失败的处理方法, 其中, 当中间域 PCE根 据第一域序列计算路径失败时还包括:
按照所述第一域序列向下游 PCE传递通知消息,使接收到所述通知消息的 下游 PCE释放掉用于进行路径计算的资源。
8. 根据权利要求 6所述的多域路径计算失败的处理方法, 其中, 还包括:
接收首域 PCE发起的针对第二域序列的路径计算请求; 其中, 所述路径计 算请求包含所述问题链路;
在进行所述第二域序列的路径计算时, 排除掉所述路径计算请求中的问题 链路。
9. 一种多域路径计算失败的处理方法, 用于尾域路径计算单元 PCE, 包括如下步 骤:
当尾域 PCE根据第一域序列计算路径失败时,将导致该计算失败的问题链 路按照所述第一域序列向上游 PCE传递。
10. 根据权利要求 9所述的多域路径计算失败的处理方法, 其中, 还包括:
接收首域 PCE发起的针对第二域序列的路径计算请求; 其中, 所述路径计 算请求包含所述问题链路;
在进行所述第二域序列的路径计算时, 排除掉所述路径计算请求中的问题 链路。
11. 一种首域路径计算单元 PCE, 包括: 问题链路确定单元, 设置为确定所有导致第一域序列发生路径计算失败的 问题链路;
删除单元, 设置为将所有问题链路从第一拓扑中删除, 得到第二拓扑; 域序列确定单元, 设置为根据所述第二拓扑确定第二域序列; 发起单元, 设置为发起针对所述第二域序列的路径计算。
12. 根据权利要求 11所述的首域路径计算单元 PCE, 其中, 所述问题链路包括: 下游 PCE的问题链路或首域 PCE的问题链路; 其中, 所述下游 PCE的问 题链是由下游 PCE传递得到的; 所述首域 PCE的问题链路是通过首域 PCE进 行路径计算得到的。
13. 根据权利要求 11 所述的首域路径计算单元 PCE, 其中, 所述域序列确定单元 包括:
第一确定子单元, 设置为根据所述第二拓扑, 确定出源节点到目的节点可 行的候选域序列;
第二确定子单元, 设置为确定所述问题链路所在的位置;
第一选取子单元, 设置为当某一中间域出现所述问题链路, 则在所述候选 域序列中排除掉包含该中间域的域序列, 之后在剩下的候选域序列中选取一个 最短的域序列作为第二域序列;若所述候选域序列均为包含该中间域的域序列, 则在该候选域序列直接选取一个最短的域序列作为第二域序列;
第二选取子单元,设置为当首域或尾域或相邻两域之间出现所述问题链路, 则在所述候选域序列中选取一最短的域序列作为第二域序列。
14. 根据权利要求 11所述的首域路径计算单元 PCE, 其中, 还包括:
第一通知单元, 设置为当首域链 PCE根据所述第一域序列计算路径失败 时, 按照所述第一域序列向下游 PCE传递通知消息, 使接收到所述通知消息的 下游 PCE释放掉用于进行路径计算的资源。
15. 根据权利要求 11 所述的首域路径计算单元 PCE, 其中, 所述发起单元具体设 置为: 按照所述第二域序列向下游 PCE发送针对该第二域序列的路径计算请求; 其中, 所述路径计算请求包含所述问题链路, 从而使得下游 PCE在针对所述第 二域序列进行路径计算时, 能够根据所述路径计算请求排除掉所述问题链路。
16. 一种中间域路径计算单元 PCE, 包括: 第一传递单元, 设置为当根据第一域序列计算路径失败时, 将导致该计算 失败的问题链路按照所述第一域序列向首域 PCE传递。
17. 根据权利要求 16所述的中间域路径计算单元 PCE, 其中, 还包括:
通知单元, 设置为当中间域 PCE根据第一域序列计算路径失败时, 按照所 述第一域序列向下游 PCE传递通知消息, 使接收到所述通知消息的下游 PCE 释放掉用于进行路径计算的资源。
18. 根据权利要求 16所述的中间域路径计算单元 PCE, 其中, 还包括:
第一接收单元,设置为接收首域 PCE发起的针对第二域序列的路径计算请 求; 其中, 所述路径计算请求包含所述问题链路;
第一排除单元, 在进行所述第二域序列的路径计算时, 排除掉所述路径计 算请求中的问题链路。
19. 一种尾域路径计算单元 PCE, 包括:
第二传递单元, 设置为当尾域 PCE根据第一域序列计算路径失败时, 将导 致该计算失败的问题链路按照所述第一域序列向上游 PCE传递。
20. 根据权利要求 19所述的尾域路径计算单元 PCE, 其中, 包括: 第二接收单元,设置为接收首域 PCE发起的针对第二域序列的路径计算请 求; 其中, 所述路径计算请求包含所述问题链路;
第二排除单元, 在进行所述第二域序列的路径计算时, 排除掉所述路径计 算请求中的问题链路。
PCT/CN2014/077313 2013-07-25 2014-05-12 一种多域路径计算失败的处理方法及路径计算单元pce WO2014183628A1 (zh)

Priority Applications (3)

Application Number Priority Date Filing Date Title
ES14797496T ES2753215T3 (es) 2013-07-25 2014-05-12 Método para procesar fallo de cálculo de ruta de múltiples zonas y elemento de cálculo de ruta (PCE)
EP14797496.8A EP3026850B1 (en) 2013-07-25 2014-05-12 Method for processing multi-zone path computation failure and path computation element (pce)
US14/907,558 US20160164715A1 (en) 2013-07-25 2014-05-12 Method for processing multi-domain path calculation failure and pce

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201310316335.XA CN104348720B (zh) 2013-07-25 2013-07-25 一种多域路径计算失败的处理方法及路径计算单元
CN201310316335.X 2013-07-25

Publications (1)

Publication Number Publication Date
WO2014183628A1 true WO2014183628A1 (zh) 2014-11-20

Family

ID=51897715

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2014/077313 WO2014183628A1 (zh) 2013-07-25 2014-05-12 一种多域路径计算失败的处理方法及路径计算单元pce

Country Status (5)

Country Link
US (1) US20160164715A1 (zh)
EP (1) EP3026850B1 (zh)
CN (1) CN104348720B (zh)
ES (1) ES2753215T3 (zh)
WO (1) WO2014183628A1 (zh)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP3032824B1 (en) * 2014-12-08 2020-08-05 ARRIS Enterprises LLC Method for automatic tracking of hybrid fiber coaxial upstream service upgrades

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103580904A (zh) * 2012-08-08 2014-02-12 中兴通讯股份有限公司 域间故障信息的发送方法及装置

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101001170A (zh) * 2006-01-10 2007-07-18 华为技术有限公司 业务故障恢复方法
CN101686197A (zh) * 2008-09-27 2010-03-31 华为技术有限公司 路径信息的获取方法和系统及路径计算装置
CN102065006A (zh) * 2010-12-01 2011-05-18 电子科技大学 一种跨域标签交换路径域间失效的恢复方法
CN102469009A (zh) * 2010-11-09 2012-05-23 中兴通讯股份有限公司 有状态路径计算单元的处理方法及有状态路径计算单元
WO2013041028A1 (en) * 2011-09-20 2013-03-28 Huawei Technologies Co., Ltd. System and Method for Computing Inter-Domain Shortest Constrained Path in a Computer Network

Family Cites Families (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
DE102005025420B4 (de) * 2005-06-02 2008-12-24 Nokia Siemens Networks Gmbh & Co.Kg Verfahren zur Bereitstellung von Ersatzwegen als schnelle Reaktion auf den Ausfall eines Links zwischen zwei Routing-Domänen
CN100446467C (zh) * 2005-07-15 2008-12-24 华为技术有限公司 链路故障的恢复方法
EP1942616B1 (en) * 2007-01-03 2014-06-25 Alcatel Lucent Method of establishing a path in a data network, network elements and data network
CN101483539B (zh) * 2008-01-07 2011-04-13 华为技术有限公司 一种获取路径的方法、路径计算单元和系统
JP4725641B2 (ja) * 2008-12-17 2011-07-13 日本テキサス・インスツルメンツ株式会社 昇降圧型スイッチングレギュレータ
CN101715149B (zh) * 2009-07-21 2013-01-23 北京邮电大学 一种多层多域分布式光网络并行跨域故障恢复方法和装置
CN101714953A (zh) * 2009-12-15 2010-05-26 中兴通讯股份有限公司 获取流量工程标签转发路径的方法和装置
US8681634B2 (en) * 2010-03-31 2014-03-25 Fujitsu Limited Systems and methods for determining protection paths in a multi-domain network

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101001170A (zh) * 2006-01-10 2007-07-18 华为技术有限公司 业务故障恢复方法
CN101686197A (zh) * 2008-09-27 2010-03-31 华为技术有限公司 路径信息的获取方法和系统及路径计算装置
CN102469009A (zh) * 2010-11-09 2012-05-23 中兴通讯股份有限公司 有状态路径计算单元的处理方法及有状态路径计算单元
CN102065006A (zh) * 2010-12-01 2011-05-18 电子科技大学 一种跨域标签交换路径域间失效的恢复方法
WO2013041028A1 (en) * 2011-09-20 2013-03-28 Huawei Technologies Co., Ltd. System and Method for Computing Inter-Domain Shortest Constrained Path in a Computer Network

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
See also references of EP3026850A4 *

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP3032824B1 (en) * 2014-12-08 2020-08-05 ARRIS Enterprises LLC Method for automatic tracking of hybrid fiber coaxial upstream service upgrades

Also Published As

Publication number Publication date
CN104348720A (zh) 2015-02-11
CN104348720B (zh) 2018-05-04
EP3026850A4 (en) 2016-08-24
ES2753215T3 (es) 2020-04-07
EP3026850A1 (en) 2016-06-01
EP3026850B1 (en) 2019-07-31
US20160164715A1 (en) 2016-06-09

Similar Documents

Publication Publication Date Title
EP2619951B1 (en) Relayed cspf computation for multiple areas and multiple autonomous systems
US9088485B2 (en) System, method and apparatus for signaling and responding to ERO expansion failure in inter-domain TE LSP
EP2625829B1 (en) System and method for computing a backup egress of a point-to-multi-point label switched path
JP5691703B2 (ja) マルチキャストネットワークシステム
CN101483539B (zh) 一种获取路径的方法、路径计算单元和系统
WO2014012207A1 (zh) 标记交换路径建立方法、数据转发方法及设备
WO2012062050A1 (zh) 有状态路径计算单元的处理方法及有状态路径计算单元
JP2013510459A (ja) 分離的なパス計算アルゴリズム
CN110086711B (zh) 流量回切方法、装置、电子设备及可读存储介质
CN105634941A (zh) 一种跨域路径的计算方法及装置
EP2997700B1 (en) Method for assured network state configuration and rollback in link-state packet networks
WO2015120602A1 (zh) 网络路径计算方法、装置和系统
US7801048B1 (en) Concurrent path computation using virtual shortest path tree
WO2014183628A1 (zh) 一种多域路径计算失败的处理方法及路径计算单元pce
US20110128888A1 (en) Distribution of routes in a network of routers
WO2018053747A1 (zh) 资源共享方法、网络节点及相关设备
WO2013135046A1 (zh) 一种建立组播路径的方法和装置
US8798050B1 (en) Re-optimization of loosely routed P2MP-TE sub-trees
US9191308B2 (en) Method and system for determining initiator and terminator in forwarding adjacency label switched path
CN103023780A (zh) 一种路由计算方法和装置
CN106411751B (zh) 一种路径优化方法及装置
JP6017036B6 (ja) インタードメインte lspにおけるero拡張障害をシグナリングする、およびそれに応答するためのシステム、方法、および装置
JP2007243482A (ja) ポイントツーマルチポイントlspに対する迂回経路算出設定方法及び装置並びにプログラム

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 14797496

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

WWE Wipo information: entry into national phase

Ref document number: 14907558

Country of ref document: US

Ref document number: 2014797496

Country of ref document: EP