WO2019237800A1 - 一种自动切分网络的方法、装置、设备及存储介质 - Google Patents

一种自动切分网络的方法、装置、设备及存储介质 Download PDF

Info

Publication number
WO2019237800A1
WO2019237800A1 PCT/CN2019/081206 CN2019081206W WO2019237800A1 WO 2019237800 A1 WO2019237800 A1 WO 2019237800A1 CN 2019081206 W CN2019081206 W CN 2019081206W WO 2019237800 A1 WO2019237800 A1 WO 2019237800A1
Authority
WO
WIPO (PCT)
Prior art keywords
network
bier
subnet
size
maximum
Prior art date
Application number
PCT/CN2019/081206
Other languages
English (en)
French (fr)
Inventor
喻敬海
高陈强
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
Publication of WO2019237800A1 publication Critical patent/WO2019237800A1/zh

Links

Images

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L45/00—Routing or path finding of packets in data switching networks
    • H04L45/02—Topology update or discovery
    • H04L45/04—Interdomain routing, e.g. hierarchical routing
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L45/00—Routing or path finding of packets in data switching networks
    • H04L45/02—Topology update or discovery
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L45/00—Routing or path finding of packets in data switching networks
    • H04L45/16—Multipoint routing
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L45/00—Routing or path finding of packets in data switching networks
    • H04L45/50—Routing or path finding of packets in data switching networks using label swapping, e.g. multi-protocol label switch [MPLS]
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L45/00—Routing or path finding of packets in data switching networks
    • H04L45/64—Routing or path finding of packets in data switching networks using an overlay routing layer

Definitions

  • the present invention relates to the technical field of network communications, and in particular, to a method, a device, a device, and a storage medium for automatically dividing a network.
  • BIER-TE Bit Index, Explicit, Replication, Traffic Engineering
  • BIER-TE based on traffic engineering, explicit replication of bit index forwarding architecture
  • the path algorithm is used to calculate the optimal path, and the packets are forwarded through the explicit hop-by-hop method. It overcomes the shortcomings that the BIER packet forwarding path is not optimal and it is difficult to adapt to the path update caused by the topology change. At the same time, it makes the network flow control more flexible.
  • multicast traffic is transmitted on the BIER-TE network.
  • a specific BIER header is required to be encapsulated on the basis of MPLS (Multi-Protocol Label Switching) packets.
  • This specific packet header It carries a BitString (bit string), and each BP (BitPosition, BIT bit) in this BitString identifies one or more adjacent points.
  • the intermediate network forwarding node forwards and copies the message according to the BitString.
  • the forwarding node looks for the BitString bit in the message and forwards the message to the corresponding one or more neighboring points. Different BitStrings need to be forwarded in different packets.
  • a BitString corresponds to a ⁇ SD, BSL, SI>, where all BPs belong to a group of ⁇ SD, BSL, SI>, and SD is a sub-domain of the BIER. Domain), BSL is the BitStringLength (bit string length) of the encapsulation header, SI is the subset identifier (Set Identifier), and the subdomain SD can be divided into multiple subsets and identified by SI.
  • the multicast stream needs to pass multiple adjacencies from the ingress node (that is, the ingress node) to each egress node (that is, the egress node).
  • the BFIR Bit-Forwarding Ingress Router, bitwise forwarding ingress router
  • the BP assigned by all neighboring points passing through a BFER must exist in a ⁇ SD, BSL, SI> at the same time, so as to ensure that the message is in a BitString Forward to that BFER.
  • the network scale is small, and the calculated network scale is less than the maximum value of BSL, which can ensure that the BPs assigned by all the neighboring points passing from BFIR to a certain BFER exist in a ⁇ SD, BSL, SI>. If the network is large and larger than the maximum maxBSL value supported by the network, no solution is given.
  • a method, an apparatus, a device, and a storage medium for automatically dividing a network can automatically divide a large-scale network in a BIER-TE-based multicast network, so that the well-divided subnets The scale is smaller than the maxBSL value, which solves the problems in BIER-TE.
  • An acquisition module for acquiring the maximum network size supported by the BIER-TE network
  • a determining module configured to determine a total network size of the BIER-TE network according to the topology information of the BIER-TE network
  • a judging module configured to determine whether the BIER-TE network meets a network segmentation condition according to the total network size and the maximum network size;
  • a segmentation module is configured to segment the BIER-TE network if the BIER-TE network meets network segmentation conditions.
  • An apparatus for automatically segmenting a network provided by an embodiment of the present invention includes a processor and a memory coupled to the processor; the memory stores an automatic segmentation network that can run on the processor.
  • a program when the program of the automatic network segmentation is executed by the processor, implements the steps of the above-mentioned method of automatic network segmentation.
  • a computer storage medium provided by an embodiment of the present invention stores a program for automatically segmenting a network, and when the program for automatically segmenting a network is executed by a processor, implements the steps of the foregoing method for automatically segmenting a network.
  • the BIER-TE network is segmented, so that the total network size of each subnet obtained by the segmentation is not greater than the maximum network size, and the message can be forwarded to the egress node of the network.
  • FIG. 1 is a schematic flowchart of a method for automatically segmenting a network according to an embodiment of the present invention
  • FIG. 2 is a schematic block diagram of an apparatus for automatically dividing a network according to an embodiment of the present invention
  • FIG. 3 is a schematic block diagram of a device for automatically dividing a network according to an embodiment of the present invention
  • FIG. 4 is a schematic diagram of a BIER-TE system architecture provided by an embodiment of the present invention.
  • FIG. 5 is a schematic diagram of a BIER-TE multicast subnet segmentation process according to an embodiment of the present invention.
  • FIG. 6 is a schematic flowchart of processing when a BIER-TE multicast egress node changes according to an embodiment of the present invention
  • FIG. 7 is a schematic flowchart of automatically dividing a network when the total number of BIER-TE networks is 455 according to an embodiment of the present invention.
  • FIG. 8 is a schematic flowchart of automatically dividing a network when the total number of BIER-TE networks is 143 according to an embodiment of the present invention
  • FIG. 9 is a schematic flowchart of automatically dividing a network when the total network size of the BIER-TE service network according to an embodiment of the present invention is 624;
  • FIG. 10 is a schematic flowchart of a process when a BIER-TE network according to an embodiment of the present invention adds 4 egress nodes.
  • FIG. 1 is a schematic flowchart of a method for automatically segmenting a network according to an embodiment of the present invention. As shown in FIG. 1, the method steps include:
  • Step S101 Obtain the maximum network size supported by the BIER-TE network.
  • the BIER-TE controller obtains the maximum network size specified by the upper layer through the northbound interface, that is, the maxBSL value (the maximum BSL).
  • the BIER-TE controller obtains, through a southbound interface, a maximum supported network size negotiated between the BIER-TE network nodes, that is, a maxBSL value (maximum BSL).
  • Step S102 Determine the total network size of the BIER-TE network according to the topology information of the BIER-TE network.
  • the topology information of the BIER-TE network includes node information (i.e., the number of nodes) and link information (i.e., the number of links).
  • the BIER-TE controller determines the number of BPs to be allocated in the BIER-TE network according to the node information and the link information, and according to the number of BPs and a preset expansion factor To determine the total network size of the BIER network. For example, the BIER-TE controller adds the number of nodes and the number of links to get the number of BPs, and then multiplies the number of BPs with the expansion factor to get the total network size.
  • Step S103 Determine whether the BIER-TE network meets a network segmentation condition according to the total network size and the maximum network size.
  • the BIER-TE controller determines that the BIER-TE network meets a network segmentation condition. For example, the total network size of the BIER-TE network is 134, and the maximum network size is 128. At this time, 134 is larger than 128, which meets the network segmentation conditions.
  • Step S104 If the BIER-TE network meets network segmentation conditions, segment the BIER-TE network.
  • the BIER-TE controller first divides the egress nodes of the BIER-TE network into N groups (for example, they are evenly divided into N groups), where N is an integer greater than 1, such as 2, 3, 4, and so on. Then, the BIER-TE network is divided into N sub-networks based on the optimal paths from the in-nodes of the BIER-TE network to the N-group out-nodes, that is, the BIER-TE controller according to the optimal path involves Nodes and links are integrated into a subnet.
  • the ingress and egress nodes of the BIER-TE network refer to the BFIR and BFER of the multicast traffic transmitted on the BIER-TE network, respectively.
  • the value of the above N is 2, that is, the outgoing nodes are divided into two groups, and then the judgment and processing are performed.
  • the BIER-TE controller after the BIER-TE controller obtains the subnet, if the subnet does not satisfy the network segmentation condition, it allocates resources to the subnet.
  • the total network size of each subnet is calculated in the same manner as in step S102. If the total network size of each of the subnets is less than or equal to the maximum network size, select no less than And the bit string length BSL that is closest to the total network size of each of the subnets, and assigns BP and the associated ⁇ SD, BSL, SI> to the nodes and links of each of the subnets.
  • the subnet needs to be segmented.
  • the total network size of each subnet is calculated in the same manner as in step S102, and then the subnets with a total network size larger than the maximum network size are segmented until the total network size of any one subnet is obtained by the segmentation. Less than or equal to the maximum network size.
  • the BIER-TE controller in this embodiment uses network segmentation to ensure that the total network size of each subnet is not greater than the maximum network size. In this way, each node and each link in the subnet can be assigned a BP.
  • the assigned BP belongs to the same ⁇ SD, BSL, SI>.
  • the BIER-TE controller After the BIER-TE controller finishes the network segmentation and allocates the BP, it can also adjust accordingly according to the change of BFER.
  • the BIER-TE controller when the BFER of the BIER-TE network is reduced, integrates the subnet from which the BFER has been deleted into a new subnet that does not include the BFER, and compares the BFER before and after the reduction of the BFER. Two subnets (ie, the subnet before the BFER reduction and the new subnet after the BFER reduction) determine the reduced nodes and links, and recover the BP allocated for the reduced nodes and links.
  • the BIER-TE controller determines the actual network of the subnet of the egress node according to the topology information of the subnet of the egress node. Scale, for example, adding the number of nodes of the subnet to which the nodes have been added and the number of links to obtain the actual network scale. If the actual network size of the subnet of the added node is greater than or equal to the maximum network size, the subnet of the added node is segmented.
  • the BIER-TE controller is based on the total network size and location of the subnet of the added node. The maximum network size is determined, and it is determined that the subnet of the added out node is divided or a new BSL is selected for the sub network of the added out node.
  • the previously selected BSL refers to the BSL selected for each subnet whose total network size is less than or equal to the maximum network size. For example, the total network size obtained for segmenting the BIER-TE network is less than or equal to the maximum network size. BSL selected for each subnet.
  • the present invention may also provide a storage medium having stored thereon a program for automatically segmenting a network.
  • a program for automatically segmentation network When the program for the automatic segmentation network is executed by a processor, the steps of the method for automatically segmenting a network described above are implemented.
  • the storage medium may include a ROM / RAM, a magnetic disk, an optical disk, and a U disk.
  • FIG. 2 is a schematic block diagram of an apparatus for automatically dividing a network according to an embodiment of the present invention. As shown in FIG. 2, the apparatus includes:
  • the obtaining module is configured to obtain the maximum network scale supported by the BIER-TE network, and implement step S101 in FIG. 1.
  • a determining module is configured to determine a total network size of the BIER-TE network according to the topology information of the BIER-TE network, and implement step S102 in FIG. 1.
  • a judging module is configured to determine whether the BIER-TE network meets a network segmentation condition according to the total network size and the maximum network size, and implement step S103 in FIG. 1.
  • a segmentation module is configured to segment the BIER-TE network if the BIER-TE network meets network segmentation conditions to implement step S104 in FIG. 1.
  • the corresponding adjustments can also be made through the cooperation of the above modules, such as re-segmenting the network, assigning BP to the newly added nodes, and recovering the former. Assigned BP and BSL, etc.
  • FIG. 3 is a schematic block diagram of a device for automatically dividing a network according to an embodiment of the present invention.
  • the device for automatically dividing a network includes a processor, and a processor coupled to the processor.
  • a memory that stores an automatic network segmentation program that can be run on the processor, and that the automatic network segmentation method is implemented by the processor when the automatic network segmentation program is executed by the processor A step of.
  • network segmentation is performed when the network scale (that is, the total network scale) is greater than the maximum maxBSL value (that is, the maximum network scale), and a suitable BSL value is calculated, so that the BPs of the same path belong to a group of ⁇ SD, BSL, SI>.
  • FIG. 4 is a schematic diagram of a BIER-TE system architecture provided by an embodiment of the present invention.
  • an upper-layer multicast stream is connected to a controller through a northbound interface (such as a RESTful API) to configure and query a multicast stream.
  • the lower-level forwarding device BFR (Bit-Forwarding Router) is connected to the controller through a southbound interface (such as Netconf / Restconf / Yang) and other protocols, collects the topology information of the device, and delivers the forwarding table and BitString.
  • src is the source and rcv is the sink receiver.
  • the network topology is collected through the southbound interface of the controller. After the network is stable, the network size is calculated. If the network size is greater than the preset maxBSL value, the network is divided into multiple subnets and divided into multiple subnets. A suitable BSL value is selected for each network. The SI of each cut network is different. The SD is the determined value configured by the upper layer to the controller, and then assigns BP to the subnet nodes and links. All BPs in the same subnet belong to A set of ⁇ SD, BSL, SI>.
  • the network topology may be a whole network topology, a user-designated topology, or a service-related topology.
  • the preset maxBSL value may be specified by an upper-layer user, or may be a supported maxBSL value negotiated between devices.
  • the method of automatically cutting the molecular net is as follows:
  • the number of nodes and links in the actual network collected by the controller through the southbound interface is M and N, respectively.
  • the number of BPs that need to be distributed throughout the network is the sum of the two, that is, M + N.
  • the total network scale S (that is, the number of BPs) can be calculated by the following formula:
  • the preset maximum supported BSL value is maxBSL. If S> maxBSL, subnet segmentation is required.
  • the controller divides the received one or more multicast egress nodes BFERs into two groups, calculates the optimal path of the two groups of ingress nodes to egress nodes, and integrates to obtain two sub-networks. All nodes and links that pass from the ingress node to the two sets of egress nodes are composed. These two sub-networks are obtained after one segmentation, and then the size of the two sub-networks is calculated as S1 and S2 using formula (1), and compared with the preset maxBSL value.
  • the BSL of the subnet is set to a value greater than or equal to and closest to S1 or S2 (BSL can only be selected from the following values: 64, 128, 256, 512, 1024, 2048, 4096); if the size of a subnet ( Assuming that the subnet S1) is still larger than the preset maxBSL value, continue to use the above method to perform network segmentation on the subnet, and iterate the segmentation until the network size is smaller than the preset maxBSL value, and select a BSL value for each segmented subnet. The BSL value is greater than or equal to and closest to the network size of this subnet.
  • FIG. 5 is a schematic diagram of a BIER-TE multicast subnet segmentation process according to an embodiment of the present invention. As shown in FIG. 5, the steps are as follows:
  • Step 201 The controller collects node information and topology of the BIER-TE network through a southbound interface.
  • Step 202 The controller accepts BIER-TE configuration related information through the northbound interface, such as Domain (a collection of forwarding devices that support BIER control protocol interaction), Sub-Domain, MPLS label range, and accepts multicast streams that are of interest to the multicast service.
  • Domain a collection of forwarding devices that support BIER control protocol interaction
  • Sub-Domain a collection of forwarding devices that support BIER control protocol interaction
  • MPLS label range accepts multicast streams that are of interest to the multicast service.
  • the ingress router (BFIR) and the egress router (BFER) the multicast service can be one or more, and the preset BSL value is obtained from the south direction or the north direction.
  • Step 203 The controller calculates the total number of nodes and links M + N according to the collected topology information, plus the expansion factor, and calculates the total network size as S using formula (1); if the network is service-related, it calculates service-related Of network nodes and links.
  • Step 204 The controller compares the calculated S value with a preset maxBSL value, and if S> maxBSL, continues processing, otherwise proceeds to step 207.
  • Step 205 The controller divides the multicast BFERs in this network into two groups, calculates the optimal paths from BFIR to the two groups of BFERs, and combines all nodes and links that the two paths pass through to form two sub-networks.
  • Step 206 The controller calculates the network scale S of the two subnets according to formula (1), and compares them with the preset maxBSL value. If S> maxBSL of a subnet, go to step 205, otherwise continue processing.
  • Step 207 The controller selects a BSL value greater than or equal to and closest to S for the subnet, and the SI value of each subnet is selected differently.
  • Step 208 The controller allocates the BP and the associated ⁇ SD, BSL, SI> of the subnet's nodes and links according to the BSL value calculated by each subnet.
  • the subnet segmentation method in this embodiment is simple. It only needs to perform 2-point segmentation on BFERs, and does not need to segment nodes on the entire network.
  • FIG. 6 is a schematic diagram of a processing flow when a BIER-TE multicast egress node changes according to an embodiment of the present invention. As shown in FIG. 6, after subnet segmentation is completed and BP is allocated, if the multicast egress node BFER changes, The allocated BP needs to be recovered for the reduced BFER and a new BP is allocated for the increased BFER. The steps are as follows:
  • Step 301 The BFER of the multicast egress node changes.
  • Step 302 Whether the number of BFERs increases, if yes, go to step 304, otherwise continue processing.
  • Step 303 Reduce the subnet corresponding to BFER, remove the reduced BFERs, recalculate the path from BFIR to the remaining BFERs, and integrate the new subnet. Compare it with the previously integrated subnet, and reduce the allocated nodes and links. BP recovery, go to step 314.
  • Step 304 Whether the user specifies to add the newly added BFERs to a certain subnet, and if a certain subnet is specified, processing continues, otherwise, go to step 313.
  • Step 305 BFERs join the user-specified subnet, recalculate the path from BFIR to the new BFERs and integrate the new subnet, and use formula (2) to calculate the new network scale S (that is, the actual network scale).
  • Step 306 Compare the calculated network scale S with maxBSL. If S ⁇ maxBSL, continue processing, otherwise, go to step 311.
  • Step 307 Whether the network scale is greater than the BSL value calculated before this subnet, if yes, continue processing; otherwise, go to step 312.
  • Step 308 Use formula (1) to recalculate the network size of the new subnet as S1 (ie, the total network size).
  • Step 309 Whether the new network scale S1 is less than maxBSL, if yes, continue processing, otherwise, go to step 311.
  • Step 310 Reselect the BSL for the subnet, re-allocate the BP for all nodes and links in the subnet, recover the previously allocated BSL and BP, and go to step 314.
  • Step 311 Use the subnet segmentation method of FIG. 5 to segment this subnet, calculate a BSL for the new subnet, and assign a BP, and then go to step 314.
  • Step 312 This subnet has enough remaining BPs to allocate BPs to the newly added nodes and links. Add a new subnet integrated by BFERs and compare it with the newly added subnets to allocate incremental nodes and links. If BP and BSL are not changed, go to step 314.
  • Step 313 Select the subnet with the smallest BSL in the network, add the new BFERs to the BFERs group of this subnet, recalculate the path from BFIR to the new BFERs group and integrate the new subnet, and use formula (2) to calculate For the new network scale S, go to step 306.
  • Step 314 There is a change in the multicast BIER, and the processing of the BP is completed.
  • formula (1) is used to calculate the network size after considering the expansion demand, that is, the total network size.
  • Formula (2) is used to calculate the network scale after BFER changes, that is, the actual network scale after BFER changes.
  • the formation of the new subnet in this embodiment is a subnet fitted based on the shortest path from BFIR to BFERs, with the least BP occupation and the shortest traffic path.
  • the BSL of this embodiment is an n-th power of 2. When so segmented, the optimal BSL is closest to maxBSL.
  • FIG. 7 is a schematic flowchart of automatic segmentation of a network when the total number of BIER-TE networks is 455 according to an embodiment of the present invention.
  • the preset maxBSL value in the BIER network is obtained from upper-level users, and the network is a network-wide topology, as shown in FIG. 7. The process is as follows:
  • Step 401 The controller collects node information and topology of the BIER network through a southbound interface, where the total number of nodes and links is 350.
  • Step 402 The controller accepts configuration BIER-related information such as Domain, Sub-Domain, and MPLS label range through the northbound interface.
  • the sub-Domain value configured by the node is 1, and the maximum supported maxBSL value is 256.
  • BFIR and BFERs are accepted.
  • Configuration, the number of BFER is 60.
  • Step 404 The network scale S value 455 is greater than the preset maxBSL value 256, and network segmentation is required.
  • Step 405 The multicast BFERs are divided into two groups, each group having 30 BFERs, and the optimal path of each group of ingress to egress is calculated.
  • the nodes and links that the two sets of paths pass through are merged into one subnet. There are two subnets in total. , Respectively for subnet 1 and subnet 2.
  • Step 406 The controller calculates the network scale S of the two subnets respectively. Assuming that the scale of subnet 1 is 260 greater than the preset maxBSL value of 256, and the scale of subnet 2 is 200 less than the preset maxBSL value of 256, the subnet 1 needs to be adjusted. Divide the network again.
  • Step 407 Divide the 30 BFERs on the subnet 1 into 2 groups again, each group of 15 and calculate the optimal path of each group of ingress to egress.
  • the nodes and links that the two groups of paths pass through are combined into one.
  • Step 408 The controller calculates the network scale S of the two subnets respectively. Assuming that the scale of subnet 3 is 190 and smaller than the preset maxBSL value of 256, the scale of subnet 4 is 120 and smaller than the preset maxBSL value of 256. Mesh segmentation.
  • Step 409 After the network is divided, three subnets are formed, which are subnet 2, subnet 3, and subnet 4.
  • the controller selects BSL values of 256, 256, and 128, and SI values of 1, 2, and 3, respectively. .
  • Step 410 The controller assigns the BSL value of each subnet to the nodes and links of the subnet and assigns the BP and the associated ⁇ SD, BSL, SI>.
  • the triples of the three subnets are ⁇ 1,256,1>, ⁇ 1,256,2> and ⁇ 1,128,3>.
  • FIG. 8 is a schematic flowchart of automatic segmentation of the network when the total number of BIER-TE networks is 143 according to an embodiment of the present invention.
  • the preset BSL value in the BIER network is the maximum maxBSL supported by negotiation between devices, and the network is the entire network.
  • the topology, as shown in Figure 8, is as follows:
  • Step 501 The controller collects node information and topology structure of the BIER network through a southbound interface, where the total number of nodes and links is 110.
  • Step 502 The controller accepts configuration BIER related information such as Domain, Sub-Domain, and MPLS label range through the northbound interface.
  • the Sub-Domain value configured by the node is 1 and the BFIR and BFERs configuration of the multicast is simultaneously received.
  • the number of BFER is 16.
  • the southbound interface accepts a maximum maxBSL value of 128 negotiated between devices.
  • Step 504 The network scale S value 143 is greater than the preset maxBSL value 128, and network segmentation is required.
  • Step 505 The multicast BFER is divided into two groups, each group of 8 BFER, and the optimal path of each group of ingress to egress is calculated. All the nodes and links that the two paths pass through are merged into one subnet, with two subnets in total. Network, respectively subnet 1 and subnet 2.
  • Step 506 The controller calculates the network scales S of the two subnets respectively. Assuming that the scale of subnet 1 is 69 and smaller than the preset maxBSL value of 128, the scale of subnet 2 is 89 and smaller than the preset maxBSL value of 128. Mesh segmentation.
  • Step 507 After the network is divided, two subnets are formed, which are subnet 1 and subnet 2, respectively.
  • the BSL values selected by the controller are 128, and the SI values are 1 and 2, respectively.
  • Step 508 The controller assigns the BSL value of each subnet to the nodes and links of the subnet and assigns BP and the associated ⁇ SD, BSL, SI>.
  • the triples of the two subnets are ⁇ 1,128,1> and ⁇ 1,128,2>.
  • FIG. 9 is a schematic flowchart of automatically dividing a network when the total network size of the BIER-TE service network according to an embodiment of the present invention is 624.
  • the BIER network is related to services. As shown in FIG. 9, the process is as follows:
  • Step 601 The controller collects node information and topology of the BIER network through a southbound interface, where the total number of nodes and links is 900.
  • Step 602 The controller accepts configuration BIER related information such as Domain, Sub-Domain, and MPLS label range through the northbound interface.
  • the Sub-Domain value configured by the node is 1, and the BFIR and BFERs configuration of the multicast is simultaneously accepted.
  • the number of BFER is 40.
  • the southbound interface accepts a maximum maxBSL value of 512 negotiated between devices.
  • Step 603 The controller calculates the optimal path according to the multicast BFIR and BFERs, and integrates the number of nodes and links of the service-related network to 480.
  • Step 604 The network scale S value 624 is greater than the preset maxBSL value 512, and network segmentation is required.
  • Step 605 The multicast BFER is divided into two groups, each group of 20 BFER, and the optimal path of each group of ingress to egress is calculated. All the nodes and links that the two groups of paths pass through are merged into one subnet. There are two subnets. Network, respectively subnet 1 and subnet 2.
  • Step 606 The controller calculates the network scales S of the two subnets respectively. Assuming that the scale of subnet 1 is 251 smaller than the preset maxBSL value of 512 and the scale of subnet 2 is 320 smaller than the preset maxBSL value of 512, no further subnetting is required. Mesh segmentation.
  • Step 607 After the network is divided, two subnets are formed, which are subnet 1 and subnet 2, respectively.
  • the BSL values selected by the controller are 256 and 512, and the SI values are 1 and 2, respectively.
  • Step 608 The controller assigns the BSL value of each subnet to the nodes and links of the subnet and assigns BP and the associated ⁇ SD, BSL, SI>.
  • the triples of the two subnets are ⁇ 1,256,1> and ⁇ 1,512,2>.
  • FIG. 10 is a schematic diagram of a processing flow when the BIER-TE network according to the embodiment of the present invention adds 4 egress nodes.
  • the BIER network has been divided into subnets according to the multicast in the embodiment of FIG. 8, BSL calculation and BP allocation have been completed.
  • the number of multicast egress nodes is increased by 4 and designated to join subnet 1.
  • the processing flow is as follows:
  • Step 601 The number of BFERs of the BIER-TE multicast is changed.
  • Step 602 Multicast adds 4 BFER.
  • Step 603 The user specifies to add the newly added BFERs to the subnet 1.
  • Step 604 Add the 4 newly added BFERs to the 8 BFERs group that has been divided into subnet 1 for a total of 12 BFERs, recalculate the path from the multicast ingress node to the 12 egress nodes, and integrate the new subnet 1 -1, use formula (2) to calculate the network scale S of the new subnet 1-1, assuming S is 70;
  • Step 605 The network scale S value 70 is less than the maxBSL value 128, and is less than the BSL value 128 calculated before the subnet 1.
  • Step 606 Subnet 1 has enough remaining BP to allocate to the newly added nodes and links.
  • the new subnet S1-1 is compared with subnet 1.
  • the extra nodes and links need to be assigned BP.
  • the link incremental allocation BP the belonging triplet is still ⁇ 1,128,1>, and the BSL is unchanged, which is 128.
  • network segmentation is performed when the total network size of the BIER-TE network is larger than the maximum network size, and a suitable BSL value is calculated, so that the BPs of the same path belong to a group of ⁇ SD, BSL, SI>, and the message is guaranteed BFER forwarded to this subnet in a BitString.

Landscapes

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

Abstract

本发明公开了一种自动切分网络的方法、装置、设备及存储介质，涉及网络通信技术领域，所述方法包括：获取基于流量工程的位索引显示复制BIER-TE网络支持的最大网络规模；根据所述BIER-TE网络的拓扑信息，确定所述BIER-TE网络的总网络规模；根据所述总网络规模和所述最大网络规模，确定所述BIER-TE网络是否满足网络切分条件；若所述BIER-TE网络满足网络切分条件，则对所述BIER-TE网络进行切分。本发明实施例通过对BIER-TE网络进行切分，使每个切分得到的子网的总网络规模不大于最大网络规模，保证报文能够转发到网络的出节点。

Description

一种自动切分网络的方法、装置、设备及存储介质
本申请要求享有2018年6月11日提交的名称为“一种自动切分网络的方法、装置、设备及存储介质”的中国专利申请CN201810593784.1的优先权，其全部内容通过引用并入本发明中。
技术领域
本发明涉及网络通信技术领域，特别涉及一种自动切分网络的方法、装置、设备及存储介质。
背景技术
BIER-TE(Bit Index Explicit Replication Traffic Engineering，BIER-TE，基于流量工程的位索引显式复制)转发架构，在BIER的基础上引入BIER-TE控制器层，并通过控制器下发转发表，同时使用路径算法计算最优路径，通过显式的逐跳方式转发报文。克服了BIER报文转发路径不是最优、难以适应拓扑变化引起的路径更新的缺点，同时使得网络流量控制更加灵活。
在组播场景下，组播流量在BIER-TE网络中传输，需要在MPLS(Multi-Protocol Label Switching，多协议标签交换)报文基础上再封装一个特定的BIER头，这个特定的报文头中携带BitString(位串)，且这个BitString中的每个BP(BitPosition，BIT位)标识一个或多个邻接点。中间网络转发节点根据BitString进行报文转发和复制。当转发节点收到组播报文时，查找报文中BitString的BIT位并将报文转发到对应的一个或多个邻接点。不同的BitString需要在不同的报文中转发，一个BitString对应于一个<SD,BSL,SI>，其中所有的BP都属于一组<SD,BSL,SI>，SD是BIER的sub-domain(子域)，BSL是封装头的BitStringLength(位串长度)，SI是子集标识(Set Identifier)，子域SD可以划分多个子集，并用SI标识。组播流从入口节点(即入节点)到每个出口节点(即出节点)需要经过多个邻接点，因此在BIER-TE中，从BFIR(Bit-Forwarding Ingress Router，按位转发入口路由器)到某一个BFER(Bit-Forwarding Egress Router，按位转发出口路由器)途中经过的所有的邻接点所分配的BP必须同时存在于一个<SD,BSL,SI>，这样才能保证报文在一个BitString中转发到该BFER。
网络规模小，计算出的网络规模小于BSL的最大值，可以保证从BFIR到某一个BFER途中经过的所有的邻接点所分配的BP都存在于一个<SD,BSL,SI>中。如果网络规模较大，并且大于网络支持的最大maxBSL值，则没有给出解决方案。
发明内容
根据本发明实施例提供的一种自动切分网络的方法、装置、设备及存储介质，在基于BIER-TE的组播网络中能够自动切分规模较大的网络，使得切分好的子网规模小于maxBSL值，解决BIER-TE中存在的问题。
本发明实施例提供的一种自动切分网络的方法，包括：
获取BIER-TE网络支持的最大网络规模；
根据所述BIER-TE网络的拓扑信息，确定所述BIER-TE网络的总网络规模；
根据所述总网络规模和所述最大网络规模，确定所述BIER-TE网络是否满足网络切分条件；
若所述BIER-TE网络满足网络切分条件，则对所述BIER-TE网络进行切分。
根据本发明实施例提供的一种自动切分网络的装置，包括：
获取模块，用于获取BIER-TE网络支持的最大网络规模；
确定模块，用于根据所述BIER-TE网络的拓扑信息，确定所述BIER-TE网络的总网络规模；
判断模块，用于根据所述总网络规模和所述最大网络规模，确定所述BIER-TE网络是否满足网络切分条件；
切分模块，用于若所述BIER-TE网络满足网络切分条件，则对所述BIER-TE网络进行切分。
本发明实施例提供的一种自动切分网络的设备包括：处理器，以及与所述处理器耦接的存储器；所述存储器上存储有可在所述处理器上运行的自动切分网络的程序，所述自动切分网络的程序被所述处理器执行时实现上述的自动切分网络的方法的步骤。
本发明实施例提供的一种计算机存储介质，存储有自动切分网络的程序，所述自动切分网络的程序被处理器执行时实现上述的自动切分网络的方法的步骤。
本发明实施例通过对BIER-TE网络进行切分，使每个切分得到的子网的总网络规模不大于最大网络规模，保证报文能够转发到网络的出节点。
附图说明
图1是本发明实施例提供的一种自动切分网络的方法的流程示意图；
图2是本发明实施例提供的一种自动切分网络的装置的示意性框图；
图3是本发明实施例提供的一种自动切分网络的设备的示意性框图；
图4是本发明实施例提供的BIER-TE系统架构示意图；
图5是本发明实施例提供的BIER-TE组播子网切分流程示意图；
图6是本发明实施例提供的BIER-TE组播出节点变化时的处理流程示意图；
图7是本发明实施例提供的BIER-TE总网络规模数为455时自动切分网络的流程示意图；
图8是本发明实施例提供的BIER-TE总网络规模数为143时自动切分网络的流程示意图；
图9是本发明实施例提供的BIER-TE的业务网的总网络规模数为624时自动切分网络的流程示意图；
图10是本发明实施例提供的BIER-TE网络增加4个出节点时的处理流程示意图。
具体实施方式
以下结合附图对本发明的主要实施例进行详细说明。
图1是本发明实施例提供的一种自动切分网络的方法的流程示意图，如图1所示，所述方法步骤包括：
步骤S101：获取BIER-TE网络支持的最大网络规模。
在一种实施方式中，BIER-TE控制器通过北向接口获取上层指定的最大网络规模，即maxBSL值(最大的BSL)。
在另一种实施方式中，BIER-TE控制器通过南向接口获取BIER-TE网络节点间协商出的共同支持的最大网络规模，即maxBSL值(最大的BSL)。
步骤S102：根据所述BIER-TE网络的拓扑信息，确定所述BIER-TE网络的总网络规模。
所述BIER-TE网络的拓扑信息包括节点信息(即节点数量)和链路信息(即链接数 量)。
在一种实施方式中，BIER-TE控制器根据所述节点信息和所述链路信息，确定所述BIER-TE网络的待分配的BP数量，并根据所述BP数量和预设的扩容因子，确定所述BIER网络的总网络规模。例如，BIER-TE控制器将节点数量和链路数量相加，得到BP数量，然后将BP数量与扩容因子相乘，得到总网络规模。
步骤S103：根据所述总网络规模和所述最大网络规模，确定所述BIER-TE网络是否满足网络切分条件。
若所述总网络规模大于所述最大网络规模，则BIER-TE控制器确定所述BIER-TE网络满足网络切分条件。例如，所述BIER-TE网络的总网络规模为134，最大网络规模为128，此时134大于128，满足网络切分条件。
步骤S104：若所述BIER-TE网络满足网络切分条件，则对所述BIER-TE网络进行切分。
BIER-TE控制器首先将所述BIER-TE网络的出节点分为N组(例如均分为N组)，N为大于1的整数，例如2、3、4等。然后根据所述BIER-TE网络的入节点分别到N组出节点的最优路径，将所述BIER-TE网络切分为N个子网，也就是说BIER-TE控制器根据最优路径涉及的节点和链路整合出子网。其中，BIER-TE网络的入节点和出节点分别指组播流量在BIER-TE网络传输时的BFIR和BFER。在一种实施方式中，上述N的取值为2，即将出节点均分为两组后进行切分判断及切分处理。
在一种实施方式中，BIER-TE控制器在得到子网后，如果子网不满足网络切分条件，则为子网分配资源。实施时，计算每个子网的总网络规模，计算方式与步骤S102相同，若每个所述子网的总网络规模小于或等于所述最大网络规模，则为每个所述子网选择不小于且最接近每个所述子网的总网络规模的位串长度BSL，并为每个所述子网的节点和链路分配BP及所属的<SD,BSL,SI>。
在一种实施方式中，BIER-TE控制器在得到子网后，如果子网满足网络切分条件，则需要对子网进行切分。实施时，计算每个子网的总网络规模，计算方式与步骤S102相同，然后对总网络规模大于所述最大网络规模的子网进行切分，直至切分得到的任意一个子网的总网络规模小于或等于所述最大网络规模。
本实施例的BIER-TE控制器通过网络切分，保证每个子网的总网络规模都不大于最大网络规模，这样，子网中的每个节点和每条链路都可以分配一个BP，其被分配的BP归属于同一<SD,BSL,SI>。
BIER-TE控制器在网络切分结束并分配好BP之后，还可以根据BFER的变化情况，进行相应调整。
在一种实施方式中，当所述BIER-TE网络的BFER减少时，BIER-TE控制器将已删除所述BFER的子网整合为不包含所述BFER的新子网，对比BFER减少前后的两个子网(即BFER减少前的子网和BFER减少后的新子网)，确定减少的节点和链路，并回收为所述减少的节点和链路分配的BP。
在一种实施方式中，当所述BIER-TE网络的出节点增加时，BIER-TE控制器根据已增加出节点的子网的拓扑信息，确定所述已增加出节点的子网的实际网络规模，例如将所述已增加出节点的子网的节点数量与链路数量相加，得到实际网络规模。若所述已增加出节点的子网的实际网络规模大于或等于所述最大网络规模，则对所述已增加出节点的子网进行切分。若所述已增加出节点的子网的实际网络规模小于所述最大网络规模且小于或等于在前选择的BSL，说明剩余BP数量足够，此时BSL不变，仅为所述已增加出节点的子网中新增的节点和链路分配BP即可。若所述已增加出节点的子网的实际网络规模小于所述最大网络规模且大于在前选择的BSL，则BIER-TE控制器根据所述已增加出节点的子网的总网络规模和所述最大网络规模，确定对所述已增加出节点的子网进行切分或为所述已增加出节点的子网选择新的BSL。其中，在前选择的BSL是指为总网络规模小于或等于最大网络规模的每个子网选择的BSL，例如为切分所述BIER-TE网络而得到的总网络规模小于或等于最大网络规模的每个子网选择的BSL。
本领域普通技术人员可以理解，实现上述实施例方法中的全部或部分步骤是可以通过程序来指令相关的硬件来完成，所述的程序可以存储于计算机可读取存储介质中。
本发明还可以提供一种存储介质，其上存储有自动切分网络的程序，所述自动切分网络的程序被处理器执行时实现上述的自动切分网络的方法的步骤。其中，所述的存储介质可以包括ROM/RAM、磁碟、光盘、U盘。
图2是本发明实施例提供的一种自动切分网络的装置的示意性框图，如图2所示，所述装置包括：
获取模块，用于获取BIER-TE网络支持的最大网络规模，实现图1步骤S101。
确定模块，用于根据所述BIER-TE网络的拓扑信息，确定所述BIER-TE网络的总网络规模，实现图1步骤S102。
判断模块，用于根据所述总网络规模和所述最大网络规模，确定所述BIER-TE网络是否满足网络切分条件，实现图1步骤S103。
切分模块，用于若所述BIER-TE网络满足网络切分条件，则对所述BIER-TE网络进行切分，实现图1步骤S104。
在一种实施例中，当BIER-TE网络的BFER增加或减少时，还可以通过上述各模块的协作，进行相应的调整，例如重新进行网络切分，为新增节点分配BP，回收在前分配的BP和BSL等。
图3是本发明实施例提供的一种自动切分网络的设备的示意性框图，如图3所示，所述一种自动切分网络的设备包括：处理器，以及与所述处理器耦接的存储器；所述存储器上存储有可在所述处理器上运行的自动切分网络的程序，所述自动切分网络的程序被所述处理器执行时实现上述的自动切分网络的方法的步骤。
本发明实施例在网络规模(即总网络规模)大于最大maxBSL值(即最大网络规模)时进行网络切分，计算出合适的BSL值，使得同一路径的BP都属于一组<SD,BSL,SI>。
图4是本发明实施例提供的BIER-TE系统架构示意图，如图4所示，上层的组播流通过北向接口(如：RESTful API)与控制器相连，进行组播流配置和查询等。下层的转发设备BFR(Bit-Forwarding Router，按位转发路由器)通过南向接口(如：Netconf/Restconf/Yang)等协议与控制器相连，收集设备的拓扑信息并下发转发表和BitString。图1中src是信源source，rcv是信宿receiver。
本发明实施例通过控制器的南向接口收集网络拓扑，待网络稳定以后，对网络规模进行计算，如果网络规模大于预先设置的maxBSL值，则要将网络切分成多个子网，并为多个子网分别选择一个合适的BSL值，每个切分子网的SI不同，SD的是上层配置给控制器的确定值，然后对子网节点和链路分配BP，同一子网的所有的BP都属于一组<SD,BSL,SI>。所述网络拓扑可以是全网拓扑、用户指定的拓扑或与业务相关的拓扑。所述预先设置的maxBSL值可以是上层用户指定的，也可以是设备间协商出来的支持的最大maxBSL值。
自动切分子网的方法如下：
控制器通过南向接口收集到的实际网络中的节点和链路数分别为M和N，则全网需要分配的BP数是两者之和，即M+N。考虑到网络扩容需求，假设扩容因子为K(一般采用推荐经验值0.3)，则总的网络规模S(即BP数)可用以下公式计算：
S＝(M+N)*(K+1)            公式(1)
预置的支持最大BSL值为maxBSL，若S>maxBSL，则需要进行子网切分。控制器将接收到的一个或多个组播的出节点BFERs均分成两组，并对两组入节点到出节点分别 进行最优路径计算，并整合得到两个子网络，这两个子网络分别由入节点到两组出节点所经过的所有节点和链路组成。这两个子网络就是一次切分后得到的，然后对两个子网络的规模分别用公式(1)进行计算为S1和S2，并与预置的maxBSL值进行比较，若子网规模小于预置maxBSL值，则将子网的BSL定为大于等于且最接近S1或S2的值(BSL只能从以下数值中选取：64、128、256、512、1024、2048、4096)；若某个子网规模(假设子网S1)还大于预置maxBSL值，继续用上述方法将子网进行网络切分，迭代切分直到网络规模小于预置maxBSL值，并为每个切分好的子网选定BSL值为大于等于且最接近此子网网络规模的BSL值。
图5是本发明实施例提供的BIER-TE组播子网切分流程示意图，如图5所示，步骤如下：
步骤201：控制器通过南向接口收集BIER-TE网络的节点信息和拓扑结构。
步骤202：控制器通过北向接口接受BIER-TE配置相关信息，如Domain(支持BIER控制协议交互的转发设备的集合)，Sub-Domain，MPLS标签范围，并接受组播业务关注的组播流的入口路由器(BFIR)和出口路由器(BFER)，组播业务可以是一个或多个，从南向或者北向获取预置的BSL值。
步骤203：控制器根据收集的拓扑信息计算节点和链路数总和M+N，加上扩容因子，总的网络规模用公式(1)计算为S；如果网络是与业务相关的则计算业务相关的网络节点和链路规模。
步骤204：控制器根据计算出的S值与预置的maxBSL值进行比较，若S>maxBSL，则继续处理，否则转步骤207。
步骤205：控制器将组播在此网络中的BFERs均分成两组，分别计算BFIR到两组BFERs的最优路径，两组路径经过的所有节点和链路合并组成两个子网络。
步骤206：控制器根据公式(1)分别计算两个子网的网络规模S，并与预置的maxBSL值进行比较，若某个子网的S>maxBSL，则转步骤205，否则继续处理。
步骤207：控制器为子网选择大于等于且最接近S的BSL值，每个子网的SI值选择不同。
步骤208：控制器根据每个子网算出的BSL值为子网的节点和链路分配BP及所属的<SD,BSL,SI>。
本实施例的子网切分方法简单，只需要对BFERs进行2分法切分，不需要对全网节 点进行切分。
图6是本发明实施例提供的BIER-TE组播出节点变化时的处理流程示意图，如图6所示，子网切分完成并分配好BP后，若组播的出节点BFER有变化，需要对减少的BFER回收已分配的BP，为增加的BFER分配新的BP，步骤如下：
步骤301：组播的的出节点BFER有变化。
步骤302：BFER个数是否增加，若是，转步骤304，否则继续处理。
步骤303：减少BFER对应的子网除去减少的BFERs，重新计算BFIR到剩余BFERs的路径并整合出新的子网，与之前整合出的子网进行对比，将减少的节点和链路所分配的BP回收，转步骤314。
步骤304：用户是否指定将新增的BFERs加入某个子网，若指定某个子网，继续处理，否则，转步骤313。
步骤305：BFERs加入用户指定子网，对BFIR到新的BFERs重新计算路径并整合出新的子网，利用公式(2)计算新的网络规模S(即实际网络规模)。
步骤306：计算出的网络规模S与maxBSL进行比较，若S<maxBSL，继续处理，否则，转步骤311。
步骤307：网络规模是否大于此子网之前计算出的BSL值，若是，继续处理，否则，转步骤312。
步骤308：使用公式(1)重新计算新子网的网络规模为S1(即总网络规模)。
步骤309：新的网络规模S1是否小于maxBSL，若是，继续处理，否则，转步骤311。
步骤310：为子网重新选择BSL，并为子网中所有节点和链路重新分配BP，回收之前分配的BSL和BP，转步骤314。
步骤311：用图5的子网切分法对此子网进行切分，为新的子网计算BSL并分配BP，转步骤314。
步骤312：此子网有足够剩余的BP给新增的节点和链路分配BP，增加BFERs整合出的新子网并与新增之前的子网比较，为增加的节点和链路增量分配BP，BSL不变，转步骤314。
步骤313：选择网络中BSL最小的子网，将新增的BFERs加入此子网的BFERs组，并对BFIR到新的BFERs组重新计算路径并整合出新的子网，利用公式(2)计算新的网络规模S，转步骤306。
步骤314：组播BIER有变化，处理BP完毕。
其中计算网络规模的公式如下：
S＝M+N                 公式(2)
可见，公式(1)用于计算考虑扩容需求后的网络规模，即总网络规模。公式(2)用于计算BFER变化后的网络规模，即BFER变化后的实际网络规模。
本实施例的新子网的形成是基于BFIR到BFERs的最短路径拟合的子网，BP占用最少，流量路径最短。
本实施例的BSL是2的n次方，当如此切分后，最优BSL最接近maxBSL。
图7是本发明实施例提供的BIER-TE总网络规模数为455时自动切分网络的流程示意图，BIER网络中预置maxBSL值从上层用户获取，网络是全网拓扑，如图7所示，其流程如下：
步骤401：控制器通过南向接口收集BIER网络的节点信息和拓扑结构，其中节点数和链路数总和为350。
步骤402：控制器通过北向接口接受配置Domain，Sub-Domain，MPLS标签范围等BIER相关信息,节点配置的Sub-Domain值为1，支持的最大maxBSL值为256；同时接受组播的BFIR和BFERs配置，BFER个数为60。
步骤403：控制器根据公式(1)算出网络规模S为350*(1+0.3)＝455。
步骤404：网络规模S值455大于预置maxBSL值256，需要进行网络切分。
步骤405：组播的BFERs均分成两组，每组30个BFER，分别计算每组入节点到出节点的最优路径，两组路径经过的节点和链接合并作为一个子网，共两个子网，分别为子网1和子网2。
步骤406：控制器分别计算两个子网的网络规模S，假设子网1的规模为260大于预置maxBSL值256，子网2的规模为200小于预置maxBSL值256，则需要对子网1再次切分网络。
步骤407：将组播在子网1的30个BFERs再次均分成2组，每组15个，分别计算每组入节点到出节点的最优路径，两组路径经过的节点和链接合并作为一个子网，共两个子网，分别为子网3和子网4。
步骤408：控制器分别计算两个子网的网络规模S，假设子网3的规模为190小于预置maxBSL值256，子网4的规模为120小于预置maxBSL值256，则不需要再进行子网 切分。
步骤409：切分网络后形成3个子网，分别是子网2，子网3和子网4，控制器为其选择的BSL值分别为256，256和128，SI值分别为1,2和3。
步骤410：控制器根据每个子网算出的BSL值为子网的节点和链路分配BP及所属的<SD,BSL,SI>，三个子网的三元组分别为<1,256,1>，<1,256,2>和<1,128,3>。
图8是本发明实施例提供的BIER-TE总网络规模数为143时自动切分网络的流程示意图，BIER网络中预置BSL值是设备间通过协商选取的支持的最大maxBSL，网络是全网拓扑，如图8所示，其流程如下：
步骤501：控制器通过南向接口收集BIER网络的节点信息和拓扑结构，其中节点数和链路数总和为110。
步骤502：控制器通过北向接口接受配置Domain，Sub-Domain，MPLS标签范围等BIER相关信息,节点配置的Sub-Domain值为1，同时接受组播的BFIR和BFERs配置，BFER个数为16，南向接口接受设备间协商的支持最大maxBSL值为128。
步骤503：控制器根据拓扑和扩容因子算出网络规模S为110*(1+0.3)＝143。
步骤504：网络规模S值143大于预置maxBSL值128，需要进行网络切分。
步骤505：组播的BFER均分成两组，每组8个BFER，分别计算每组入节点到出节点的最优路径，两组路径经过的所有节点和链接合并作为一个子网，共两个子网，分别为子网1和子网2。
步骤506：控制器分别计算两个子网的网络规模S，假设子网1的规模为69小于预置maxBSL值128，子网2的规模为89小于预置maxBSL值128，则不需要再进行子网切分。
步骤507：切分网络后形成2个子网，分别是子网1和子网2，控制器为其选择的BSL值都为128，SI值分别为1和2。
步骤508：控制器根据每个子网算出的BSL值为子网的节点和链路分配BP及所属的<SD,BSL,SI>，两个子网的三元组分别为<1,128,1>和<1,128,2>。
图9是本发明实施例提供的BIER-TE的业务网的总网络规模数为624时自动切分网络的流程示意图，BIER网络与业务相关，如图9所示，其流程如下：
步骤601：控制器通过南向接口收集BIER网络的节点信息和拓扑结构，其中节点数和链路数总和为900。
步骤602：控制器通过北向接口接受配置Domain，Sub-Domain，MPLS标签范围等BIER相关信息,节点配置的Sub-Domain值为1，同时接受组播的BFIR和BFERs配置，BFER个数为40，南向接口接受设备间协商的支持最大maxBSL值为512。
步骤603：控制器根据组播BFIR和BFERs计算最优路径，并整合出业务相关网络的节点和链接数为480，根据公式(1)算出业务网络规模S为480*(1+0.3)＝624。
步骤604：网络规模S值624大于预置maxBSL值512，需要进行网络切分。
步骤605：组播的BFER均分成两组，每组20个BFER，分别计算每组入节点到出节点的最优路径，两组路径经过的所有节点和链接合并作为一个子网，共两个子网，分别为子网1和子网2。
步骤606：控制器分别计算两个子网的网络规模S，假设子网1的规模为251小于预置maxBSL值512，子网2的规模为320小于预置maxBSL值512，则不需要再进行子网切分。
步骤607：切分网络后形成2个子网，分别是子网1和子网2，控制器为其选择的BSL值分别为256和512，SI值分别为1和2。
步骤608：控制器根据每个子网算出的BSL值为子网的节点和链路分配BP及所属的<SD,BSL,SI>，两个子网的三元组分别为<1,256,1>和<1,512,2>。
图10是本发明实施例提供的BIER-TE网络增加4个出节点时的处理流程示意图，BIER网络已按图8实施例的组播切分好子网，BSL计算和BP分配都已完成，组播出节点个数增加4个，并指定加入子网1中，如图10所示，处理流程如下：
步骤601：BIER-TE组播的BFERs个数有变化。
步骤602：组播增加4个BFER。
步骤603：用户指定将新增的BFERs加入子网1中。
步骤604：新增的4个BFER加入子网1已切分的8个BFERs组中，共12个BFER，重新计算组播入节点到12个出节点的路径，并整合出新的子网1-1，用公式(2)计算新子网1-1的网络规模S，假设S为70；
步骤605：网络规模S值70小于maxBSL值128，且小于子网1之前计算的BSL值128；
步骤606：子网1有足够剩余的BP给新增的节点和链路分配，新子网S1-1与子网1进行比较，多出的节点和链路是需要分配BP的，为这些节点和链路增量分配BP，所属 三元组还是<1,128,1>，且BSL不变，为128。
综上所述，本发明实施例具有以下技术效果：
本发明实施例在BIER-TE网络的总网络规模大于最大网络规模时进行网络切分，计算出合适的BSL值，使得同一路径的BP都属于一组<SD,BSL,SI>，保证报文在一个BitString中转发到该子网的BFER。
尽管上文对本发明进行了详细说明，但是本发明不限于此，本技术领域技术人员可以根据本发明的原理进行各种修改。因此，凡按照本发明原理所作的修改，都应当理解为落入本发明的保护范围。

Claims (11)

  1. 一种自动切分网络的方法，其中，包括：
    获取基于流量工程的位索引显示复制BIER-TE网络支持的最大网络规模；
    根据所述BIER-TE网络的拓扑信息，确定所述BIER-TE网络的总网络规模；
    根据所述总网络规模和所述最大网络规模，确定所述BIER-TE网络是否满足网络切分条件；
    若所述BIER-TE网络满足网络切分条件，则对所述BIER-TE网络进行切分。
  2. 根据权利要求1所述的方法，其中，所述获取BIER-TE网络支持的最大网络规模的步骤包括：
    通过北向接口获取上层指定的最大网络规模，或者通过南向接口获取节点间协商的最大网络规模。
  3. 根据权利要求2所述的方法，其中，所述BIER-TE网络的拓扑信息包括节点信息和链路信息，所述根据所述BIER-TE网络的拓扑信息，确定所述BIER-TE网络的总网络规模包括：
    根据所述节点信息和所述链路信息，确定所述BIER-TE网络的待分配的比特位BP数量；
    根据所述BP数量和预设的扩容因子，确定所述BIER网络的总网络规模。
  4. 根据权利要求3所述的方法，其中，所述根据所述总网络规模和所述最大网络规模，确定所述BIER-TE网络是否满足网络切分条件包括：
    若所述总网络规模大于所述最大网络规模，则确定所述BIER-TE网络满足网络切分条件。
  5. 根据权利要求4所述的方法，其中，所述若所述BIER-TE网络满足网络切分条件，则对所述BIER-TE网络进行切分包括：
    将所述BIER-TE网络的出节点分为N组，N为大于1的整数；
    根据所述BIER-TE网络的入节点分别到N组出节点的最优路径，将所述BIER-TE网络切分为N个子网；
    对总网络规模大于所述最大网络规模的子网进行切分，直至切分得到的任意一个子网的总网络规模小于或等于所述最大网络规模；
    若每个所述子网的总网络规模小于或等于所述最大网络规模，则为每个所述子网选择不小于且最接近每个所述子网的总网络规模的位串长度BSL，并为每个所述子网的节点和链路分配BP。
  6. 根据权利要求5所述的方法，其中，所述N的取值为2。
  7. 根据权利要求2-6任意一项所述的方法，其中，在所述若所述BIER-TE网络满足网络切分条件，则对所述BIER-TE网络进行切分之后，还包括：
    当所述BIER-TE网络的出节点减少时，将已删除所述出节点的子网整合为不包含所述出节点的新子网；
    对比所述出节点减少前的子网和所述出节点减少后得到的新子网，确定减少的节点和链路，并回收为所述减少的节点和链路分配的BP。
  8. 根据权利要求2-6任意一项所述的方法，其中，在所述若所述BIER-TE网络满足网络切分条件，则对所述BIER-TE网络进行切分之后，还包括：
    当所述BIER-TE网络的出节点增加时，根据已增加出节点的子网的拓扑信息，确定所述已增加出节点的子网的实际网络规模；
    若所述已增加出节点的子网的实际网络规模大于或等于所述最大网络规模，则对所述已增加出节点的子网进行切分；
    若所述已增加出节点的子网的实际网络规模小于所述最大网络规模且小于或等于在前选择的所述BSL，则为所述已增加出节点的子网中新增的节点和链路分配BP；
    若所述已增加出节点的子网的实际网络规模小于所述最大网络规模且大于在前选择的所述BSL，则根据所述已增加出节点的子网的总网络规模和所述最大网络规模，确定对所述已增加出节点的子网进行切分或为所述已增加出节点的子网选择新的BSL。
  9. 一种自动切分网络的装置，其中，包括：
    获取模块，用于获取基于流量工程的位索引显示复制BIER-TE网络支持的最大网络规模；
    确定模块，用于根据所述BIER-TE网络的拓扑信息，确定所述BIER-TE网络的总网络规模；
    判断模块，用于根据所述总网络规模和所述最大网络规模，确定所述BIER-TE网络是否满足网络切分条件；
    切分模块，用于若所述BIER-TE网络满足网络切分条件，则对所述BIER-TE网络进 行切分。
  10. 一种自动切分网络的设备，其中，所述设备包括：处理器，以及与所述处理器耦接的存储器；所述存储器上存储有可在所述处理器上运行的自动切分网络的程序，所述自动切分网络的程序被所述处理器执行时实现如权利要求1至8中任一项所述的自动切分网络的方法的步骤。
  11. 一种计算机存储介质，其中，存储有自动切分网络的程序，所述自动切分网络的程序被处理器执行时实现如权利要求1至8中任一项所述的自动切分网络的方法的步骤。
PCT/CN2019/081206 2018-06-11 2019-04-03 一种自动切分网络的方法、装置、设备及存储介质 WO2019237800A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201810593784.1 2018-06-11
CN201810593784.1A CN110581806B (zh) 2018-06-11 2018-06-11 一种自动切分网络的方法、装置、设备及存储介质

Publications (1)

Publication Number Publication Date
WO2019237800A1 true WO2019237800A1 (zh) 2019-12-19

Family

ID=68809440

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2019/081206 WO2019237800A1 (zh) 2018-06-11 2019-04-03 一种自动切分网络的方法、装置、设备及存储介质

Country Status (2)

Country Link
CN (1) CN110581806B (zh)
WO (1) WO2019237800A1 (zh)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115022241A (zh) * 2022-05-31 2022-09-06 烽火通信科技股份有限公司 一种bier自动配置及管理bsl的方法和装置
WO2022204651A1 (en) * 2021-03-22 2022-09-29 Cisco Technology, Inc. Efficient bier forwarding over varying bsl domain using header stitching

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112511319B (zh) * 2020-07-03 2024-06-28 中兴通讯股份有限公司 信息通告、报文转发、报文消除方法、装置、设备和介质

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101945033A (zh) * 2009-07-08 2011-01-12 中兴通讯股份有限公司 一种流量工程域合并/分割时的处理方法、装置及系统
CN107623630A (zh) * 2016-07-13 2018-01-23 中兴通讯股份有限公司 一种位索引显式复制信息传递方法和装置
US20180063024A1 (en) * 2016-08-30 2018-03-01 Cisco Technology, Inc. Transparent and efficient multi-destination tcp communications based on bit indexed explicit replication
CN107872385A (zh) * 2017-10-11 2018-04-03 中国电子科技集团公司第三十研究所 一种sdn网络路由计算与控制方法

Family Cites Families (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US9544230B2 (en) * 2013-09-17 2017-01-10 Cisco Technology, Inc. Migration support for bit indexed explicit replication
EP3047604B1 (en) * 2013-09-17 2021-03-24 Cisco Technology, Inc. Bit indexed explicit replication
US11451474B2 (en) * 2013-09-17 2022-09-20 Cisco Technology, Inc. Equal cost multi-path with bit indexed explicit replication
US9432205B2 (en) * 2014-11-04 2016-08-30 Telefonaktiebolaget L M Ericsson (Publ) Explicit block encoding of multicast group membership information with bit index explicit replication (BIER)
CN106341327A (zh) * 2015-07-08 2017-01-18 中兴通讯股份有限公司 一种bier报文的传输方法及系统
CN107770070B (zh) * 2016-08-17 2022-07-08 中兴通讯股份有限公司 一种信息传输的方法、设备和系统
CN107968750B (zh) * 2016-10-20 2021-06-15 中兴通讯股份有限公司 报文传输方法、装置及节点
US10637675B2 (en) * 2016-11-09 2020-04-28 Cisco Technology, Inc. Area-specific broadcasting using bit indexed explicit replication

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101945033A (zh) * 2009-07-08 2011-01-12 中兴通讯股份有限公司 一种流量工程域合并/分割时的处理方法、装置及系统
CN107623630A (zh) * 2016-07-13 2018-01-23 中兴通讯股份有限公司 一种位索引显式复制信息传递方法和装置
US20180063024A1 (en) * 2016-08-30 2018-03-01 Cisco Technology, Inc. Transparent and efficient multi-destination tcp communications based on bit indexed explicit replication
CN107872385A (zh) * 2017-10-11 2018-04-03 中国电子科技集团公司第三十研究所 一种sdn网络路由计算与控制方法

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2022204651A1 (en) * 2021-03-22 2022-09-29 Cisco Technology, Inc. Efficient bier forwarding over varying bsl domain using header stitching
US11705983B2 (en) 2021-03-22 2023-07-18 Cisco Technology, Inc. Efficient BIER forwarding over varying BSL domain using header stitching
CN115022241A (zh) * 2022-05-31 2022-09-06 烽火通信科技股份有限公司 一种bier自动配置及管理bsl的方法和装置
CN115022241B (zh) * 2022-05-31 2023-06-09 烽火通信科技股份有限公司 一种bier自动配置及管理bsl的方法和装置

Also Published As

Publication number Publication date
CN110581806A (zh) 2019-12-17
CN110581806B (zh) 2022-02-25

Similar Documents

Publication Publication Date Title
CN105871722B (zh) 标签构造以及标签报文转发方法及装置
JP7162117B2 (ja) ネットワークスライス制御方法及び装置、コンピュータ読み取り可能な記憶媒体
EP3208977B1 (en) Data forwarding method, device and system in software-defined networking
CN111492627B (zh) 为不同应用建立不同隧道的基于控制器的服务策略映射
KR101503629B1 (ko) 주소 기반 캐리어 네트워크의 구별 전달
CN111654438B (zh) 一种网络中建立转发路径的方法、控制器及系统
CN101019382B (zh) 通过环形拓扑实现多点服务
WO2019237800A1 (zh) 一种自动切分网络的方法、装置、设备及存储介质
EP3043517A1 (en) Routing method, device and system
EP1388979B1 (en) Method and apparatus for end-to-end connection between an RPR and an MPLS network
US20090034984A1 (en) Sonet method and system having network service access point addressing
JP7121759B2 (ja) 通信方法、ネットワークサーバ、及びコンピュータ読取可能記憶媒体
JP2002141932A (ja) ネットワーク共有帯域割当て方法及びこれを用いるネットワークシステム
CN103477612A (zh) 经扩展以连接网络层级的云服务控制和管理架构
EP3836492B1 (en) Pseudo wire load sharing method and device
EP3176987B1 (en) Communication control device, communication control method and communication system
US8897295B2 (en) Method and system for providing traffic engineering interworking
WO2015024440A1 (zh) 一种获取ip链路的链路开销值的方法及系统
CN101945048A (zh) 一种标签交换路径的建立方法、系统及装置
EP1592182A1 (en) Data transmission device and data transmission system
WO2004028043A1 (fr) Procede de distribution d&#39;itineraires pour l&#39;optimisation de services multiples dans un reseau de transmission sdh
WO2012068834A1 (zh) 聚合链路的管理方法及系统
CN104702510B (zh) 一种rpr网络中的报文转发方法及装置
CN101488900A (zh) 在下一代网络中传送业务的实现方法及系统
CN101577932B (zh) 在下一代网络中传送业务的实现方法及系统

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: 19818509

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

32PN Ep: public notification in the ep bulletin as address of the adressee cannot be established

Free format text: NOTING OF LOSS OF RIGHTS PURSUANT TO RULE 112(1) EPC ( EPO FORM 1205A DATED 20/04/2021 )

122 Ep: pct application non-entry in european phase

Ref document number: 19818509

Country of ref document: EP

Kind code of ref document: A1