[go: up one dir, main page]

WO2018233580A1 - 一种网络中建立转发路径的方法、控制器及系统 - Google Patents

一种网络中建立转发路径的方法、控制器及系统 Download PDF

Info

Publication number
WO2018233580A1
WO2018233580A1 PCT/CN2018/091688 CN2018091688W WO2018233580A1 WO 2018233580 A1 WO2018233580 A1 WO 2018233580A1 CN 2018091688 W CN2018091688 W CN 2018091688W WO 2018233580 A1 WO2018233580 A1 WO 2018233580A1
Authority
WO
WIPO (PCT)
Prior art keywords
delay
service
forwarding path
dcs
ratio
Prior art date
Legal status (The legal status 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 status listed.)
Ceased
Application number
PCT/CN2018/091688
Other languages
English (en)
French (fr)
Inventor
黄宗和
林程勇
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Huawei Technologies Co Ltd
Original Assignee
Huawei Technologies Co Ltd
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 Huawei Technologies Co Ltd filed Critical Huawei Technologies Co Ltd
Priority to EP18821178.3A priority Critical patent/EP3618370B1/en
Publication of WO2018233580A1 publication Critical patent/WO2018233580A1/zh
Priority to US16/703,675 priority patent/US11290568B2/en
Anticipated expiration legal-status Critical
Ceased legal-status Critical Current

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
    • H04L45/12—Shortest path evaluation
    • H04L45/121—Shortest path evaluation by minimising delays
    • 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
    • H04L45/30—Routing of multiclass traffic
    • 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
    • H04L45/302—Route determination based on requested QoS
    • 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
    • H04L45/302—Route determination based on requested QoS
    • H04L45/306—Route determination based on the nature of the carried application
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L67/00—Network arrangements or protocols for supporting network services or applications
    • H04L67/01—Protocols
    • H04L67/10—Protocols in which an application is distributed across nodes in the network
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L67/00—Network arrangements or protocols for supporting network services or applications
    • H04L67/50—Network services
    • H04L67/60—Scheduling or organising the servicing of application requests, e.g. requests for application data transmissions using the analysis and optimisation of the required network resources
    • H04L67/63—Routing a service request depending on the request content or context
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L12/00—Data switching networks
    • H04L12/28—Data switching networks characterised by path configuration, e.g. LAN [Local Area Networks] or WAN [Wide Area Networks]
    • H04L12/46—Interconnection of networks
    • H04L12/4641—Virtual LANs, VLANs, e.g. virtual private networks [VPN]
    • 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
    • H04L45/12—Shortest path evaluation
    • H04L45/125—Shortest path evaluation based on throughput or bandwidth
    • 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
    • H04L69/00—Network arrangements, protocols or services independent of the application payload and not provided for in the other groups of this subclass
    • H04L69/18—Multiprotocol handlers, e.g. single devices capable of handling multiple protocols

Definitions

  • the present invention relates to the field of communications, and in particular, to a method, a controller, and a system for establishing a forwarding path in a network.
  • an end-to-end user data transmission passes through a network interface, such as an NI1 core router (Provider Edge, PE), such as PE1, and an interconnect router (Operator Gateway Border Router).
  • a network interface such as an NI1 core router (Provider Edge, PE), such as PE1, and an interconnect router (Operator Gateway Border Router).
  • OGBR Interconnect Router
  • ASs Autonomous Systems
  • the layered controller is evolved based on the architecture of the domain controller (DC).
  • a super controller (SC) is set up on the upper layer of the DC.
  • One SC can jointly manage multiple domain controls.
  • the tiered controllers are deployed to solve the problem of multi-protocol label switching (MPLS) virtual private network (VPN) service deployment scenarios because DC cannot implement cross-domain Business deployment.
  • MPLS multi-protocol label switching
  • VPN virtual private network
  • SC can manage DCs of different vendors, thus simplifying network operation and maintenance.
  • SDN Software Defined Network
  • LOD Latency On Demand
  • E2E end-to-end
  • VPLS virtual private LAN service
  • the present application provides a method for establishing a forwarding path in a network, which is used to solve the technical problem that the delay constraint of the forwarding path cannot be allocated for different services in the prior art.
  • an embodiment of the present application provides a method for establishing a forwarding path in a network, where the network includes an SC and multiple DCs, where the SC manages the multiple DCs, and each of the multiple DCs respectively Manage a domain, the method includes:
  • the SC receives a service request message, where the service request message includes a service type and a first time delay, and the service request message requests the SC to establish a forwarding path for transmitting the service corresponding to the service type in the network,
  • the delay of the forwarding path is less than or equal to the first delay
  • the SC selects a delay allocation manner according to the service request message
  • the first delay into a plurality of delay segments, where the plurality of delay segments are in one-to-one correspondence with the plurality of DCs;
  • each of the plurality of delay segments Sending, by the SC, each of the plurality of delay segments to a DC corresponding to each of the delay segments, triggering each of the plurality of DCs to be in a domain managed by the SC Establishing a forwarding path segment, where the delay of each of the forwarding path segments is less than or equal to a delay segment corresponding to each DC;
  • the SC connects the plurality of forwarding path segments established by the multiple DCs to form the forwarding path.
  • the delay of the established forwarding path can be customized and intelligently matched according to the service requirement, and the service type is judged, and then the distribution mode suitable for the service is selected, thereby solving the problem that there is no suitable principle in the prior art.
  • the above technical solution can make the network utilization more optimal when the service path is iterated, and make the network optimization more satisfy the customer's demand.
  • the time delay allocation manner of the SC selection is average Decomposing the first delay into a plurality of delay segments according to the ALRD mode, in an ALV mode:
  • the SC averages the first delay into the plurality of delay segments according to the number of the forwarding path segments.
  • the time delay allocation mode selected by the SC is a Minimum Latency Rate Distribution (MIRD) mode.
  • MIRD Minimum Latency Rate Distribution
  • the SC decomposes the first time delay into the plurality of time delay segments according to a ratio between the obtained plurality of the second time delays.
  • the time delay allocation mode selected by the SC is Maximum Latency Rate Distribution (MARD).
  • MARD Maximum Latency Rate Distribution
  • the SC decomposes the first time delay into the plurality of time delay segments according to a ratio between the obtained plurality of the third time delays.
  • the SC selects the delay allocation mode as an average fiber utilization rate distribution (Average Fiber Utilization Rate Distribution).
  • the SC decomposes the first delay into multiple delay segments according to the AFRD mode, including:
  • the SC decomposes the first time delay into the plurality of time delay segments according to a ratio between obtaining the plurality of fiber utilization rates.
  • the AFRD mode can reflect the link usage in the domain, and the network utilization is better when the service path is iterated.
  • the time delay allocation mode selected by the SC is a total bandwidth ratio ratio allocation (BORD) mode.
  • Decomposing the first delay into a plurality of delay segments according to the BORD manner includes:
  • the SC obtains traffic in a domain managed by each of the plurality of DCs, and calculates, according to the obtained traffic, a ratio of traffic of each DC to a total bandwidth of each DC;
  • the SC decomposes the first time delay into the plurality of time delay segments according to a ratio between a ratio of obtaining a traffic of each of the DCs to a total bandwidth of each of the DCs.
  • the BORD mode can reflect the link usage in the domain, and the network utilization is better when the service path is iterated.
  • the slot allocation calculation manner of the SC selection is a specified delay ratio allocation manner, wherein the SC decomposes the first delay into the multiple delay segments according to a ratio specified by a user. .
  • the SC can select different delay ratio allocation methods for different service types, so that the delay constraint of the forwarding path can be dynamically allocated to meet different service requirements.
  • an embodiment of the present application provides an SC that establishes a forwarding path in a network, where the network includes the SC and multiple domain controllers DC, and the SC manages the multiple DCs, and the multiple DCs Each DC in the management manages a domain, and the SC includes:
  • a receiving unit configured to receive a service request message, where the service request message includes a service type and a first time delay, where the service request message requests the SC to establish, in the network, a forwarding for transmitting a service corresponding to the service type a path, where a delay of the forwarding path is less than or equal to the first delay;
  • a selecting unit configured to select a delay allocation manner according to the service request message
  • a processing unit configured to decompose the first delay into a plurality of delay segments according to the delay allocation manner, where the plurality of delay segments are in one-to-one correspondence with the plurality of DCs;
  • a sending unit configured to separately send each of the plurality of delay segments to a DC corresponding to each of the delay segments, and trigger each of the plurality of DCs to be managed by the DC Establishing a forwarding path segment in the domain, where the delay of each of the forwarding path segments is less than or equal to a delay segment corresponding to each DC;
  • the processing module is further configured to connect the multiple forwarding path segments established by the multiple DCs to form the forwarding path.
  • the SC further includes:
  • a determining unit configured to determine a number of network nodes in the domain in which the forwarding path segment is established
  • the selecting unit configured to: when the determining unit determines that a difference between the number of network nodes in the domain in which the forwarding path segment is established is within a predetermined range, the selected delay allocation mode is an ALRD mode, among them
  • the first time delay is averagely decomposed into the plurality of delay segments according to the number of the forwarding path segments.
  • the SC further includes:
  • a determining unit configured to determine that the service identified by the service type is a service sensitive to delay
  • the selecting unit is configured to: when the determining unit determines that the service identified by the service type is a service that is sensitive to delay, the selected delay allocation mode is an MIRD mode, and the SC is determined according to the MIRD mode.
  • Decomposing the first delay into a plurality of delay segments includes:
  • the SC decomposes the first time delay into the plurality of time delay segments according to a ratio between the obtained plurality of the second time delays.
  • the SC further includes:
  • a determining unit configured to determine that the service identified by the service type is a service that is not sensitive to delay
  • the selecting unit is configured to: when the determining unit determines that the service identified by the service type is a service that is not sensitive to delay, the selected delay allocation mode is a MARD mode, and the SC is according to the MARD mode.
  • Decomposing the first delay into a plurality of delay segments includes:
  • the SC decomposes the first time delay into the plurality of time delay segments according to a ratio between the obtained plurality of the third time delays.
  • the SC further includes:
  • a determining unit configured to determine that the service identified by the service type is a service whose transmission file size exceeds a preset value
  • the selecting unit is configured to: when the determining unit determines that the service of the service type identifier is a service whose transmission file size exceeds a preset value, the selected delay allocation mode is an AFRD mode, and the SC is according to the Decomposing the first delay into multiple delay segments in the AFRD mode includes:
  • the SC decomposes the first time delay into the plurality of time delay segments according to a ratio between obtaining the plurality of fiber utilization rates.
  • the SC further includes:
  • a determining unit configured to determine that the service identified by the service type is a big data service
  • the selecting unit is configured to: when the determining unit determines that the service of the service type identifier is a big data service, the selected delay allocation mode is a BORD mode, and the SC is configured according to the BORD manner. Decomposing a delay into multiple delay segments includes:
  • the SC obtains traffic in a domain managed by each of the plurality of DCs, and calculates, according to the obtained traffic, a ratio of traffic of each DC to a total bandwidth of each DC;
  • the SC decomposes the first time delay into the plurality of time delay segments according to a ratio between a ratio of obtaining a traffic of each of the DCs to a total bandwidth of each of the DCs.
  • the delay allocation mode selected by the selecting unit is a specified delay ratio mode, wherein the delay is decomposed into the plurality of delay segments according to a ratio specified by a user.
  • an embodiment of the present application provides a method for establishing a forwarding path in a network, where the network includes an SC and multiple DCs, and the SC manages the multiple DCs, and each of the multiple DCs Managing a domain separately, the method includes:
  • the DC establishes a forwarding path segment in the managed domain, where the delay of the forwarding path segment is less than or equal to the delay segment;
  • the DC sends the information of the forwarding path segment to the SC, and triggers the SC to connect multiple forwarding path segments established by the multiple DCs to form the forwarding path.
  • the delay of the established forwarding path can be customized and intelligently matched according to the service requirement, and the service type is judged, and then the distribution mode suitable for the service is selected, thereby solving the problem that there is no suitable principle in the prior art.
  • the above solution can make the network utilization better when the service path is iterated, and make the network optimization more satisfy the customer's demand.
  • the delay selection mode of the SC selection is an ALRD mode.
  • Decomposing the first delay into the delay period according to the ALRD mode includes:
  • the SC averages the first delay into the delay period according to the number of the DCs.
  • the time delay allocation mode selected by the SC is an MIRD mode, and the SC is configured according to the MIRD mode. Decomposing the first delay into the delay period includes:
  • the SC decomposes the first time delay into the time delay segment according to a ratio between the obtained plurality of the second time delays.
  • the time delay allocation mode selected by the SC is a MARD mode
  • the SC according to the MARD mode Decomposing the first time delay into the time delay segment includes:
  • the SC decomposes the first time delay into the time delay segment according to a ratio between the obtained plurality of the third time delays.
  • the time delay allocation mode selected by the SC is an AFRD mode, and the SC is according to the AFRD mode.
  • Decomposing the first delay into the delay period includes:
  • the SC decomposes the first time delay into the time delay segment according to a ratio between obtaining the plurality of fiber utilization rates.
  • the AFRD mode can reflect the link usage in the domain, and the network utilization is better when the service path is iterated.
  • the time delay allocation mode selected by the SC is a BORD mode, and the SC performs the first time according to the BORD mode. Deferring into the delay period includes:
  • the SC obtains traffic in a domain managed by each of the plurality of DCs, and calculates, according to the obtained traffic, a ratio of traffic of each DC to a total bandwidth of each DC;
  • the SC decomposes the first delay into the delay period according to a ratio between a ratio of obtaining a plurality of the flows of each DC to a total bandwidth of each DC.
  • the BORD mode can reflect the link usage in the domain, and the network utilization is better when the service path is iterated.
  • the slot allocation calculation manner of the SC selection is a specified delay ratio allocation manner, wherein the SC decomposes the first delay into the delay period according to a ratio specified by a user. .
  • an embodiment of the present application provides a DC that establishes a forwarding path in a network, where the network includes a super controller SC and multiple DCs, and the SC manages the multiple DCs, and the multiple DCs Each DC in the management manages one domain separately, and the DC includes:
  • a receiving unit configured to receive a delay segment sent by the SC, where the delay segment is formed by decomposing the first delay in the service request message after the SC selects a delay allocation manner according to the service request message.
  • a establishing unit configured to establish a forwarding path segment in the domain, where a delay of the forwarding path segment is less than or equal to the delay segment;
  • a sending unit configured to send the information of the forwarding path segment to the SC, and trigger the SC to connect the multiple forwarding path segments established by the establishing unit to form the forwarding path.
  • the embodiment of the present application provides a network system for establishing a forwarding path, including an SC, a first DC, and a second DC, where the SC manages the first DC and the second DC, where the first The DC and the second DC respectively manage a domain,
  • the SC is configured to receive a service request message, where the service request message includes a service type and a first time delay, where the service request message is used to request the SC to establish, in the network, a transport type corresponding to the service type.
  • a forwarding path of the service where the delay of the forwarding path is less than or equal to the first delay, and the delay allocation mode is selected according to the service request message, and the first delay is decomposed into the delay allocation manner according to the delay allocation manner.
  • the first time delay segment and the second time delay segment are sent to the first DC, and the second time delay segment is sent to the second DC;
  • the first DC is configured to receive the first delay segment sent by the SC, and establish a first forwarding path segment in the domain that it manages according to the first delay segment, where the first forwarding path segment The delay of the first forwarding path is less than or equal to the first delay period, and the information of the first forwarding path is sent to the SC;
  • the second DC is configured to receive the second delay segment sent by the SC, and establish a second forwarding path segment in the domain that it manages according to the second delay segment, where the second forwarding path segment The delay is less than or equal to the second delay period, and the information of the second forwarding path is sent to the SC;
  • the SC is further configured to connect the first forwarding path segment and the second forwarding path segment to form the forwarding path according to the information of the first forwarding path and the information of the second forwarding path.
  • an embodiment of the present application provides a network device, where the network device includes: a processor, a network interface, and a memory.
  • the memory can be used to store the program code
  • the processor is used to call the program instructions in the memory to perform the method designed in the foregoing aspects.
  • specific execution steps refer to the foregoing aspects, and details are not described herein again.
  • the embodiment of the present application provides a computer storage medium for storing computer software instructions used by the network device, which includes a program designed to perform the above aspects.
  • FIG. 1 is a schematic structural diagram of an end-to-end network in the prior art according to an embodiment of the present invention
  • FIG. 2 is a schematic diagram of an application network scenario of a method for establishing a forwarding path according to an embodiment of the present invention
  • FIG. 3 is a flowchart of a method for establishing a forwarding path according to an embodiment of the present invention
  • FIG. 4 is a schematic diagram of an application scenario of a method for establishing a forwarding path according to an embodiment of the present invention
  • FIG. 5 is a schematic diagram of a delay allocation method corresponding to different service types according to an embodiment of the present invention.
  • FIG. 6 is a schematic diagram of an application scenario of a method for establishing a forwarding path according to an embodiment of the present invention
  • FIG. 6B is a schematic diagram of an application scenario of a method for establishing a forwarding path according to an embodiment of the present invention
  • FIG. 7 is a schematic structural diagram of a controller for establishing a forwarding path according to an embodiment of the present invention.
  • FIG. 8 is a flowchart of a method for establishing a forwarding path according to an embodiment of the present invention.
  • FIG. 9 is a schematic structural diagram of a controller for establishing a forwarding path according to an embodiment of the present invention.
  • FIG. 10 is a schematic diagram of a network system for establishing a forwarding path according to an embodiment of the present invention.
  • FIG. 11 is a schematic structural diagram of a network device according to an embodiment of the present invention.
  • FIG. 2 is a schematic diagram of an application network scenario of a method for establishing a forwarding path in a network according to an embodiment of the present disclosure.
  • FIG. 2 shows a network system composed of a physical network 1, a super controller 2, a single domain controller 3, a collaborator 4, and a Utraffic 5.
  • the coordinator is used for E2E service delivery and interface presentation; the functions of the SC include: E2E service decomposition, decomposition into a single domain configuration model; network resource management, management network resources (Route-Distinguisher, RD), RT (Route-Target) , PW-ID (Password-Identification), etc.; topology management, network-wide physical topology, L3 topology, and service topology state management; tunnel management, E2E label switched path (LSP) planning management; DC function Including: network element management, physical network element discovery, inventory information acquisition, etc.; network element resource management, management network element resources; topology management, intra-domain physical topology discovery, intra-domain L3 topology discovery, status update, etc.; tunnel management, intra-domain TE tunnel management Path calculation, TE path calculation in the domain; Utraffic 5 is used for traffic monitoring and delivery tuning.
  • network resource management management network resources (Route-Distinguisher, RD), RT (Route-Target)
  • SDN Software Defined Network
  • LOD Latency On Demand
  • VPLS virtual private LAN service
  • Performance metrics for a single carrier network end-to-end performance metrics * (2/3 + 1/3 * 0.35) * 1/2.
  • Performance index of the inter-network interconnection link end-to-end performance indicator *1/3*0.65.
  • the embodiment of the present application provides a method for establishing a forwarding path in a network.
  • a physical network including three domains is taken as an example, and the following SDN cross-domain network is supported to support a method for releasing a VPN service with E2E delay constraint.
  • 3 is a flow chart of a method of establishing a forwarding path in a network in accordance with a first embodiment of the present invention.
  • the method includes:
  • the SC receives a service request message, where the service request message includes a service type and a first time delay, where the service request message is used to request the SC to establish, in the network, a forwarding for transmitting the service corresponding to the service type. a path, where a delay of the forwarding path is less than or equal to the first delay;
  • the SC selects a delay allocation manner according to the service request message.
  • the SC decomposes the first time delay into multiple time delay segments according to the time delay allocation manner, where the multiple time delay segments are in one-to-one correspondence with the multiple DCs.
  • the SC sends each of the multiple delay segments to a DC corresponding to each of the delay segments, and triggers each DC of the multiple DCs in a domain that is managed by the SC. Establishing a forwarding path segment, where the delay of each of the forwarding path segments is less than or equal to a delay segment corresponding to each DC;
  • the SC connects the multiple forwarding path segments established by the multiple DCs to form the forwarding path.
  • the SC connects multiple forwarding path segments to form a forwarding path.
  • the LSPs in each domain are connected together to form an entire forwarding path by means of LSP splicing.
  • the delay selection mode of the SC selection is an ALRD mode.
  • Decomposing the first delay into multiple delay segments according to the ALRD mode includes:
  • the SC averages the first delay into the plurality of delay segments according to the number of the forwarding path segments.
  • the time delay allocation mode selected by the SC is an MIRD mode, and the SC is configured according to the MIRD mode.
  • Decomposing the first delay into a plurality of delay segments includes:
  • the SC decomposes the first time delay into the plurality of time delay segments according to a ratio between the obtained plurality of the second time delays.
  • the time delay allocation mode selected by the SC is a MARD mode
  • the SC according to the MARD mode Decomposing the first delay into a plurality of delay segments includes:
  • the SC decomposes the first time delay into the plurality of time delay segments according to a ratio between the obtained plurality of the third time delays.
  • the time delay allocation mode selected by the SC is an AFRD mode, and the SC is according to the AFRD mode.
  • Decomposing the first delay into a plurality of delay segments includes:
  • the SC decomposes the first time delay into the plurality of time delay segments according to a ratio between obtaining the plurality of fiber utilization rates.
  • the time delay allocation mode selected by the SC is a BORD mode, and the SC performs the first time according to the BORD mode.
  • Decomposing into multiple delay segments includes:
  • the SC obtains traffic in a domain managed by each of the plurality of DCs, and calculates, according to the obtained traffic, a ratio of traffic of each DC to a total bandwidth of each DC;
  • the SC decomposes the first delay into the plurality of delay segments according to a ratio between a ratio of obtaining a plurality of the flows of each of the DCs to a total bandwidth of each of the DCs.
  • the entire network includes SC and 3 DCs and 3 AS domains managed by 3 DCs (SC and DC are not shown).
  • the information of each domain is shown in Table 1:
  • Td1 Node(pe1)+node(abr1)+node(asbr1)+link(pe1-abr1)+link(abr1-asbr1)
  • Td2 node(pe1)+node(abr1)+node(abr2)+node(asbr2)+node(asbr1)+link(pe1-abr1)+lin k(abr1-abr2)+link(abr2-asbr2)+link (asbr2-asbr1)
  • Table 2 is the domain link delay attributes calculated according to the above formula.
  • the following is an example of the case where the L3VPN service of the three ASs is deployed between PE1 and PE3, and the delay requirement is Tsvc.
  • the steps for issuing the L3VPN service are as follows:
  • the northbound application delivers services with delay constraints through the message.
  • Service packets are sent to the SC controller through the restconf interface.
  • the SC control restserver receives the northbound request and then forwards the northbound request to the L3 Service Management (SM) module.
  • SM Service Management
  • the L3SM module converts the Internet Engineering Task Force (IETF) model into a stable NetL3VPN model.
  • the NetL3vpn on the SC interprets the northbound packet and applies for network-level resources RD and RT. If the "latency" carried in the northbound packet is extracted, the service feature is extracted, and the POLICY module is called according to the service feature to obtain the delay allocation method.
  • the other type can be pre-configured through the Web User Interface (WEBUI) interface.
  • WEBUI Web User Interface
  • the northbound APP can specify the service type to be created as the specified service type.
  • FIG. 5 is a schematic diagram of different allocation methods corresponding to different service types according to an embodiment of the present invention.
  • Ti is the delay allocated by the i-th domain
  • TTotal is the total delay constraint of the tunnel
  • DTotal is the sum of the autonomous domains of the tunnel.
  • the delay constraint of each AS domain is 3.3ms according to the average delay.
  • This type of allocation is applicable to scenarios where the network size of the autonomous domain across the tunnel is comparable. For example, the difference in the number of network nodes is within a predetermined range, such as 10%. For example, when a core router is connected to two metropolitan area networks of equal size, and a tunnel with a delay of 10 ms needs to be created between the two metropolitan area networks, the allocation method can be used, for example, for Internet Protocol Security (Internet Protocol Security, IPSEC) and other services.
  • Internet Protocol Security Internet Protocol Security
  • the delay ratio specified by the user Assigned according to the delay ratio specified by the user. If the user specifies 2:3:5, the total delay is 10ms.
  • the delay allocated by each domain is 2ms, 3ms, and 5ms.
  • This type of allocation is generally a scenario where the network administrator combines his own network planning and delay planning, and at the same time, according to the network load monitoring situation, the real-time online delay allocation method. If the network administrator finds that a certain domain has a large load and a small domain load is small, the network administrator can manually adjust the delay allocation ratio and customize the delay allocation.
  • T Min Min(Delay j )
  • TMin is the delay value of the path with the smallest delay among all the paths in the domain
  • Delayj is the delay of the jth path.
  • the minimum delay ratio of each AS domain is 1:2:2
  • the total delay is 10ms
  • the delays allocated by each domain are 2ms, 4ms, and 4ms respectively.
  • This allocation method the delay of the domain allocation with small delay is small, the delay of allocation delay with large delay is large, and the delay of light load network and heavy load network is considered.
  • This method is a relatively simple allocation method and can reach The network load is tuned.
  • This allocation applies to services that are sensitive to latency constraints, such as L3VPN, L2VPN, and EVPN.
  • the delay-sensitive service specifically refers to that the delay for transmitting the service cannot be greater than a specific threshold, and the specific threshold may be specified by a user according to actual conditions.
  • T MAX Max(Delay j )
  • TMAX is the delay value of the path with the largest delay among all the paths in the domain
  • Delayj is the delay of the jth path.
  • the delay allocated by each domain is 4ms, 2ms, and 4ms.
  • This method is a more flexible allocation method, and the maximum delay of the domain allocation with large delay is more, so that there are more paths in the domain. It is suitable for services with relatively low latency sensitivity, such as INTERNET, FTP, TELNET, EMAIL or P2P.
  • FURi fiber utilization rate
  • the delay of each domain allocation is 4ms, 3ms, 3ms.
  • the average fiber utilization calculation method (fiber 1 through the flow / fiber 1 maximum bandwidth + ... + fiber n through the flow / fiber n maximum bandwidth) / n * 100%.
  • This allocation method is allocated from the intra-domain fiber utilization ratio, which is beneficial to improving the bandwidth utilization in the domain and can be used for trans-domain services for transferring large files.
  • the so-called large files are files whose size exceeds a preset value.
  • Cross-domain services for transmitting images and/or sounds such as video (VIDEO) and Voice over Internet Protocol (VOIP).
  • BORI bandwidth occupied rate
  • Flowj the traffic passing the j-th link
  • BandWidthj the maximum physical bandwidth of the j-th link.
  • the delay of each domain allocation is 1ms, 3ms, 6ms.
  • the calculation method of the total traffic ratio in the domain (the traffic through link 1 +... + the traffic through link n) / (the maximum bandwidth of link 1 + ... + the maximum bandwidth of link n) * 100%.
  • This allocation method treats the domain as a black box, and the ratio of total traffic to total capacity is similar to the average fiber utilization in the domain.
  • the allocation method is considered from the perspective of network bandwidth utilization. Therefore, when used for general data services, especially for big data (Big Data), such as VXLAN, this allocation method is better than "average fiber utilization in the domain".
  • the average fiber utilization ratio and total traffic ratio method can reflect the link usage in the domain. These two methods can make the network utilization better when the service path is iterated.
  • FIG. 5 is a correspondence diagram of a delay allocation method and a specific service type according to an embodiment of the present invention.
  • the allocation method of "minimum delay ratio" should be selected.
  • the required time delay Tsvc will be allocated in accordance with the ratio of Td1:Td9:Td17. Therefore, the delays allocated for each domain are Tsvc*Td1/(Td1+Td9+Td17), Tsvc*Td9/(Td1+Td9+Td17), and Tsvc*Td17/(Td1+Td9+Td17).
  • the establishment of the communication link is performed after the step of allocating the delay constraints of the respective domains according to the selected allocation method.
  • the NetL3vpn on the SC divides the model service based on the DC granularity, and decomposes it into domain-level network data and sends it to the DC controller.
  • the DC receives the network data from the SC, invokes the SNC component, and dynamically creates the TE tunnel as the bearer tunnel of the L3VPN based on the delay requirement, that is, creates a forwarding path with a delay less than or equal to the allocated delay segment.
  • For the first domain dynamically create a TE (Traffic Engineering) tunnel according to the delay constraint of Tsvc*Td1/(Td1+Td9+Td17) ms; for the second domain, according to Tsvc*Td9/(Td1) +Td9+Td17) ms delay constraint to dynamically create a TE tunnel; for the third domain, dynamically create a TE tunnel according to the delay constraint of Tsvc*Td17/(Td1+Td9+Td17) ms.
  • TE Traffic Engineering
  • the DC sends the information about the TE tunnel to the SC, triggering the SC to perform LSP splicing, and connecting the TE tunnels in each domain to form an entire forwarding path.
  • the tunnel delay actually iterated to follow the principle of minimum delay preference is Tsvc*Td1/(Td1+Td9+Td17)ms, Tsvc*Td9/(Td1+Td9+Td17)ms, Tsvc*Td17/(Td1+Td9+Td17 ) ms, the tunnel delay actually iterated to follow the principle of minimum delay preference. Note that the tunnel iterated here is a two-way shared tunnel.
  • Figure 6A shows the same physical network as Figure 4, containing three AS domains. The information of each domain is shown in Table 3:
  • an L3VPN service that spans three AS domains is deployed between PE1 and PE3.
  • the delay is 20 ms.
  • the minimum delay ratios in each domain are calculated to be 2.45 ms, 2.55 ms, and 0.9 ms, respectively, using the minimum delay ratio allocation method.
  • each domain allocates a delay according to a 2.45:2.55:0.9 ratio, and the delays allocated by each domain are 8.3 ms, 8.7 ms, and 3 ms, respectively.
  • the SC sends the three delay segments to the corresponding DCs respectively, thereby triggering the DC to establish a forwarding path, that is, a tunnel, in the domain it manages.
  • For the first domain dynamically create a TE tunnel according to the 8.3ms delay constraint; for the second domain, dynamically create a TE tunnel according to the 8.7ms delay constraint; for the third domain, according to the 3ms time Delay constraint to dynamically create a TE tunnel.
  • the tunnel constraint has a delay constraint, and the tunnel creation mode adopts an exact matching method. That is, the tunnel iteration management module selects a path that exactly matches the delay requirement according to the delay constraint condition. If not found, create a new tunnel that meets the delay extension. To create a tunnel with a delay constraint of 50ms, the PCE module calculates the path according to the constraint conditions. The delay of the path is the linear sum of the node and the link delay on the path. If there are multiple paths that satisfy the delay requirement, you can select the delay accumulation and the smallest one.
  • the three domain iterative tunnels are shown as thick solid lines 61, 62 and 63 in Figure 6B. Since the delay constraints of the iterative tunnel are 8.3 ms, 8.7 ms, and 3 ms, respectively, the delay of the forwarding path created by each DC needs to be less than or equal to the delay constraint segment allocated by each. In actual use, the iterative tunnel delay can follow the principle of minimum delay preference.
  • the delay of the tunnel 61 that is finally created is 2.45 ms
  • the delay of the tunnel 62 is 2.55 ms
  • the delay of the tunnel 63 is 0.9 ms.
  • the present invention also provides a method of establishing a forwarding path in another network, wherein the network includes the SC and the plurality of DCs as described above.
  • the method includes: the DC receives a delay period sent by the SC, where the delay period is that the SC selects a delay allocation manner according to the service request message, and then decomposes the first delay in the service request message. ;
  • the DC establishes a forwarding path segment in the domain, where the delay of the forwarding path segment is less than or equal to the delay segment;
  • the DC sends the information of the forwarding path segment to the SC, and triggers the SC to connect the plurality of forwarding path segments to form the forwarding path.
  • the method according to the first embodiment of the present invention can also monitor whether the delay allocation of each domain satisfies the service requirement by monitoring the network.
  • the delay parameter is modified, and then the delay constraint of each domain is reallocated. If the determination is yes, the current time domain constraints of each domain are maintained.
  • the method provided by the embodiment of the present invention supports the online modification function of the user.
  • the input of the user may be introduced, thereby modifying the delay parameter value online.
  • the tunnel in the controller domain will be iterated according to the new delay to satisfy the delay constraint. After the iteration is successful, the VPN will switch to the new path.
  • the UTRAffic 5 monitors network link utilization, delay, jitter, and packet loss ratio to adjust the delay allocation method.
  • the DC controller obtains real-time network performance data such as single-domain average fiber utilization, single-domain total traffic ratio, link delay, and node delay through the interface provided by the Utraffic 5. Therefore, the user monitors the E2E service delay indicator through the Utraffic 5, and when it finds that the service is not satisfied, the controller can trigger the dynamic online modification of the link delay, trigger the recurrence of the service path, and re-identify and allocate the delay of each domain. Assignment, so that the business traffic path is adjusted to a better path.
  • the user's online input can be introduced. Users can modify the service delay parameters online, and introduce artificially customized delay constraint modification, so that network optimization can better meet customer needs.
  • the existing technology node and link delay attributes are static configurations.
  • the preset delay method is strong in real time.
  • the network performance data such as the minimum delay and maximum delay, fiber utilization, node delay, and link delay can be obtained through the interface provided by the monitor (such as Utraffic) and refresh the controller node and link attributes in real time or periodically. Real-time, can reflect the real state of the network for a certain period of time.
  • an embodiment of the present application further provides an SC700 for establishing a forwarding path in a network.
  • the SC includes:
  • the receiving unit 701 is configured to receive a service request message, where the service request message includes a service type and a first time delay, and the service request message requests the SC to establish, in the network, a service for transmitting the service type corresponding service. Forwarding a path, where a delay of the forwarding path is less than or equal to the first delay;
  • the selecting unit 702 is configured to select a delay allocation manner according to the service request message.
  • the processing unit 703 is configured to decompose the first delay into a plurality of delay segments according to the delay allocation manner, where the multiple delay segments are in one-to-one correspondence with the multiple DCs;
  • the sending unit 704 is configured to separately send each of the multiple delay segments to a DC corresponding to each of the delay segments, and trigger each of the multiple DCs to manage a forwarding path segment is established in the domain, and the delay of each of the forwarding path segments is less than or equal to a delay segment corresponding to each DC;
  • the processing unit 703 is further configured to connect the multiple forwarding path segments established by the multiple DCs to form the forwarding path.
  • the SC700 according to the embodiment of the present application further includes:
  • a determining unit configured to determine a number of network nodes in the domain in which the forwarding path segment is established
  • the selecting unit is configured to: when the determining unit determines that a difference between the number of network nodes in the domain in which the forwarding path segment is established is within a predetermined range, the selected delay allocation mode is an ALRD mode;
  • Decomposing the first delay into a plurality of delay segments according to the delay allocation manner includes:
  • the processing unit averages the first delay into the plurality of delay segments according to the number of the plurality of delay segments.
  • the SC700 according to the embodiment of the present application further includes:
  • a determining unit configured to determine whether the service identified by the service type is a service sensitive to delay
  • the selecting unit is configured to: when the determining unit determines that the service identified by the service type is a service that is sensitive to delay, the selected delay allocation mode is an MIRD mode;
  • Decomposing the first delay into a plurality of delay segments according to the MIRD manner includes:
  • the processing unit obtains a second delay of a path with the smallest delay among the paths in the domain managed by each of the plurality of DCs;
  • the processing unit decomposes the first time delay into the plurality of time delay segments according to a ratio between the obtained plurality of the second time delays.
  • the SC700 according to the embodiment of the present application further includes:
  • a determining unit configured to determine whether the service identified by the service type is a service that is not sensitive to delay
  • the selecting unit is configured to: when the determining unit determines that the service identified by the service type is a service that is not sensitive to delay, the selected delay allocation mode is a MARD mode;
  • Decomposing the first delay into the plurality of delay segments according to the MARD mode includes:
  • the processing unit obtains a third delay of a path with the largest delay among the paths in the domain managed by each of the plurality of DCs;
  • the processing unit decomposes the first time delay into the plurality of time delay segments according to a ratio between the obtained plurality of the third time delays.
  • the SC700 according to the embodiment of the present application further includes:
  • a determining unit configured to determine whether the service identified by the service type is a service whose transmission file size exceeds a preset value
  • the selecting unit is configured to: when the determining unit determines that the service of the service type identifier is a service whose transmission file size exceeds a preset value, the selected delay allocation mode is an AFRD mode;
  • Decomposing the first delay into a plurality of delay segments according to the AFRD manner includes:
  • the processing unit obtains a fiber utilization rate in a domain managed by each of the plurality of DCs;
  • the processing unit decomposes the first time delay into the plurality of time delay segments according to a ratio between obtaining the plurality of fiber utilization rates.
  • the SC700 according to the embodiment of the present application further includes:
  • a determining unit configured to determine whether the service identified by the service type is a big data service
  • the selecting unit is configured to: when the determining unit determines that the service identified by the service type is a big data service, the selected delay allocation mode is a BORD mode;
  • Decomposing the first delay into the plurality of delay segments according to the BORD manner includes:
  • the processing unit obtains traffic in a domain managed by each of the plurality of DCs, and calculates, according to the obtained traffic, a ratio of traffic of each DC to a total bandwidth of each DC;
  • the processing unit decomposes the first time delay into the plurality of time delay segments according to a ratio between a ratio of obtaining a plurality of the flows of each of the DCs to a total bandwidth of each of the DCs.
  • the delay allocation mode selected by the selecting unit is a specified delay ratio mode, wherein the processing unit decomposes the first delay into the plurality according to a ratio specified by a user. Delay period.
  • the network includes a super controller SC and a plurality of domain controllers DC, and the SC manages the plurality of DCs, the plurality of Each DC in the DC manages one domain separately.
  • the method includes:
  • the DC receives a delay segment sent by the SC, where the delay segment is formed by decomposing the first delay in the service request message after the SC selects a delay allocation manner according to the service request message. ;
  • the DC establishes a forwarding path segment in the managed domain, where a delay of the forwarding path segment is less than or equal to the delay segment.
  • the DC sends the information of the forwarding path segment to the SC, and triggers the SC to connect the multiple forwarding path segments established by the multiple DCs to form the forwarding path.
  • the delay selection mode of the SC selection is an ALRD mode.
  • Decomposing the first delay into the delay period according to the ALRD mode includes:
  • the SC averages the first delay into the delay period according to the number of the DCs.
  • the time delay allocation mode selected by the SC is an MIRD mode, and the SC is configured according to the MIRD mode. Decomposing the first delay into the delay period includes:
  • the SC decomposes the first time delay into the time delay segment according to a ratio between the obtained plurality of the second time delays.
  • the time delay allocation mode selected by the SC is a MARD mode
  • the SC according to the MARD mode Decomposing the first time delay into the time delay segment includes:
  • the SC decomposes the first time delay into the time delay segment according to a ratio between the obtained plurality of the third time delays.
  • the time delay allocation mode selected by the SC is an AFRD mode, and the SC is according to the AFRD mode.
  • Decomposing the first delay into the delay period includes:
  • the SC decomposes the first time delay into the time delay segment according to a ratio between obtaining the plurality of fiber utilization rates.
  • the time delay allocation mode selected by the SC is a BORD mode, and the SC performs the first time according to the BORD mode. Deferring into the delay period includes:
  • the SC obtains traffic in a domain managed by each of the plurality of DCs, and calculates, according to the obtained traffic, a ratio of traffic of each DC to a total bandwidth of each DC;
  • the SC decomposes the first delay into the delay period according to a ratio between a ratio of obtaining a plurality of the flows of each DC to a total bandwidth of each DC.
  • the slot allocation calculation manner of the SC selection is a specified delay ratio allocation manner, wherein the SC decomposes the first delay into the delay period according to a ratio specified by a user. .
  • an embodiment of the present application further provides a domain controller DC that establishes a forwarding path in a network, where the network includes a super controller SC and a plurality of the DCs, and the SC manages the multiple DCs.
  • the DC 900 includes:
  • the receiving unit 901 is configured to receive a delay segment sent by the SC, where the SC delays the first time delay in the service request message after the SC selects a delay allocation manner according to the service request message. to make;
  • the establishing unit 902 is configured to establish a forwarding path segment in the domain, where a delay of the forwarding path segment is less than or equal to the delay segment;
  • the sending unit 903 is configured to send the information of the forwarding path segment to the SC, and trigger the SC to connect the plurality of the forwarding path segments established by the establishing unit to form the forwarding path.
  • a network system 1000 for establishing a forwarding path is also provided.
  • the system 1000 includes an SC1001, a first DC 1002, and a second DC 1003.
  • the SC 1001 manages a first DC 1002 and a second DC 1003.
  • the first DC 1002 and the second DC 1003 respectively manage a domain, where
  • the SC1001 is configured to receive a service request message, where the service request message includes a service type and a first time delay, where the service request message is used to request the SC 1001 to establish, in the network, a forwarding path for transmitting the service corresponding to the service type.
  • the delay of the forwarding path is less than or equal to the first delay, and the delay allocation mode is selected according to the service request message, and the first delay is decomposed into the first according to the delay allocation manner.
  • a time delay segment and the second time delay segment, and the first time delay segment is sent to the first DC 1002, and the second time delay segment is sent to the second DC 1003;
  • the first DC 1002 is configured to receive the first delay segment sent by the SC1001, and establish a first forwarding path segment in the domain that is managed by the first time delay segment, where the delay of the first forwarding path segment is less than Equal to the first delay segment, sending information of the first forwarding path to the SC1001;
  • a second DC 1003 configured to receive the second delay segment sent by the SC1001, and establish a second forwarding path segment in the domain that is managed by the second time delay segment, where the delay of the second forwarding path segment is less than Equal to the second delay period, sending information of the second forwarding path to the SC1001;
  • the SC1001 is further configured to connect the first forwarding path segment and the second forwarding path segment to form the forwarding path according to the information of the first forwarding path and the information of the second forwarding path.
  • the embodiment of the present application further provides a network device.
  • the network device 1100 can perform the method shown in the foregoing embodiment.
  • the network device 1100 includes a processor 1101, a network interface 1102, and a memory 1103.
  • the processor 1101 includes, but is not limited to, a central processing unit (CPU), a network processor (NP), an application-specific integrated circuit (ASIC), or a programmable logic device (Programmable Logic Device, One or more of the PLDs.
  • the PLD may be a Complex Programmable Logic Device (CPLD), a Field-Programmable Gate Array (FPGA), or a Generic Array Logic (GAL).
  • CPLD Complex Programmable Logic Device
  • FPGA Field-Programmable Gate Array
  • GAL Generic Array Logic
  • the memory 1103 includes, but is not limited to, a Content-Addressable Memory (CAM), such as a Ternary CAM (TCAM) or a Random-Access Memory (RAM).
  • CAM Content-Addressable Memory
  • TCAM Ternary CAM
  • RAM Random-Access Memory
  • the network interface 1102 can include a common physical interface, and the physical interface can be an Ethernet interface or an Asynchronous Transfer Mode (ATM) interface.
  • the processor 1101, the network interface 1102, and the memory 1103 can be integrated into one or more separate circuits or hardware, such as an ASIC.
  • the memory 1103 can be used to store the program code and data of the network device 1100.
  • the processor 1101 is used to invoke the program instructions in the memory 1103 to execute the method in the embodiment shown in FIG. 3. For the specific implementation steps, refer to the foregoing embodiment, where No longer.
  • a fourth embodiment of the present invention provides a computer storage medium for storing computer software instructions for use in the network device, including a program designed to perform the method of establishing a forwarding path.
  • computer readable media includes both computer storage media and communication media including any medium that facilitates transfer of a computer program from one location to another.
  • the storage medium may be any available medium that can be accessed by a general purpose or special purpose computer.
  • the storage medium may be at least one of the following: a read-only memory (ROM), a RAM, a magnetic disk, or an optical disk.
  • ROM read-only memory
  • RAM random access memory
  • magnetic disk a magnetic disk
  • optical disk optical disk
  • the method and controller provided in accordance with embodiments of the present invention enable delay support for customizable and intelligent matching, solving the technical problem of not being able to allocate delay constraints for forwarding paths for different services.
  • the above embodiments are all described based on spanning VPN services, in practice, the methods and controllers of the present invention can be used for other types of cross-domain services as well.
  • the functions described herein can be implemented in hardware, software, firmware, or any combination thereof.
  • the functions may be stored in a computer readable medium or transmitted as one or more instructions or code on a computer readable medium.
  • Computer readable media includes both computer storage media and communication media including any medium that facilitates transfer of a computer program from one location to another.
  • the storage medium may be any available medium that can be accessed by a general purpose or special purpose computer.
  • the storage medium may be at least one of the following: a read-only memory (ROM), a RAM, a magnetic disk, or an optical disk.
  • ROM read-only memory
  • RAM random access memory
  • magnetic disk a magnetic disk
  • optical disk optical disk

Landscapes

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

Abstract

本发明涉及通信领域，特别是涉及网络中建立转发路径的方法，该网络包括超级控制器SC和多个域控制器DC，其中SC管理多个DC，多个DC中的每个DC分别管理一个域，该方法包括SC接收业务请求消息，SC根据所述业务请求消息选择时延分配方式；SC根据所述时延分配方式将所述第一时延分解成多个时延段，SC分别将所述多个时延段中的每个时延段发送到与所述每个时延段对应的DC，触发所述多个DC中的每个DC在其管理的域中建立转发路径段，以及SC将多个DC建立的多个所述转发路径段连接形成转发路径。另外，本发明还涉及网络中建立转发路径的控制器以及建立转发路径的系统。

Description

一种网络中建立转发路径的方法、控制器及系统
本申请要求于2017年6月20日提交中国专利局、申请号为201710470662.9、申请名称为“一种网络中建立转发路径的方法、控制器及系统”的中国专利申请的优先权，其全部内容通过引用结合在本申请中。
技术领域
本发明涉及通信领域，特别是涉及一种网络中建立转发路径的方法、控制器及系统。
背景技术
如图1所示，一个端到端的用户数据传输经过省网的汇聚路由器(Network Interface,NI)，如：NI1核心路由器(Provider Edge，PE)，如PE1、互连路由器(Operator Gateway Border Router，OGBR)进入另外一个省网的互连路由器、核心路由器直至目的用户的终端设备。该数据传输跨越两个自治域(Autonomous System，AS)，如：AS-1和AS-2。
分层控制器是在域控制器(Domain Controller，DC)的架构基础上演进出来的，是在DC的上层再架设一个超级控制器(Super Controller，SC)，一个SC可以协同管理多个域控制器。分层控制器的出现，主要是为了解决跨域无缝多协议标签交换(Multi-Protocol Label Switching，MPLS)虚拟专用网络(Virtual Private Network，VPN)业务布放场景，因为DC无法实现跨域的业务布放。另一方面，对运营商而言，SC可以管理不同厂商的DC，从而简化网络的运维。
而目前软件定义网络(Software Defined Network，SDN)跨域业务按需时延(Latency On Demand，LOD)尚未支持带端到端(End to End，E2E)时延约束的业务发放。假设在一个跨3个AS域的网络，用户需要部署一个10ms时延要求的虚拟专用局域网业务(Virtual Private LAN Service，VPLS)专线业务，到底是按照3：3：4分配时延还是按照2：5：3分配，或者当网络拓扑或者容量发生变化时，该如何分配，这些都是尚待解决的技术问题。
发明内容
本申请提供了一种网络中建立转发路径的方法，用于解决现有技术中无法针对不同的业务来分配转发路径的时延约束的技术问题。
第一方面，本申请实施例提供了一种网络中建立转发路径的方法，该网络包括SC和多个DC，所述SC管理所述多个DC，所述多个DC中的每个DC分别管理一个域，所述方法包括：
所述SC接收业务请求消息，所述业务请求消息包含业务类型和第一时延，所述业务请求消息请求所述SC在所述网络中建立用于传输所述业务类型对应业务的转发路径，所述转发路径的时延小于或等于所述第一时延；
所述SC根据所述业务请求消息选择时延分配方式；
所述SC根据所述时延分配方式将所述第一时延分解成多个时延段，所述多个时延段与所述多个DC一一对应；
所述SC分别将所述多个时延段中的每个时延段发送到与所述每个时延段对应的DC，触发所述多个DC中的每个DC在其管理的域中建立转发路径段，每个所述转发路径段的时延小于或等于与所述每个DC对应的时延段；
所述SC将所述多个DC建立的多个转发路径段连接形成所述转发路径。
在上述技术方案中，所建立的转发路径的时延可以根据业务需求进行定制和智能匹配，通过对业务类型进行判断，进而选择适合此业务的分配方式，解决了现有技术中没有合适的原则来分配转发路径的时延约束的技术问题，上述技术方案能够使业务路径迭代时网络利用率更优，以及使得网络优化更能满足客户的需求。
在一个可能的设计中，当所述多个DC中的每个DC所管理的域内的网络节点的数量的差值在预定范围内时，所述SC选择的所述时延分配方式为平均时延分配(Average Latency Rate Distribution，ALRD)方式，所述SC根据所述ALRD方式将所述第一时延分解成多个时延段包括：
所述SC按照所述转发路径段的个数将所述第一时延平均分解成所述多个时延段。
在一个可能的设计中，当所述业务类型标识的业务为对时延敏感的业务时，所述SC选择的所述时延分配方式为最小时延比率分配(Minimum Latency Rate Distribution，MIRD)方式，所述SC根据所述MIRD方式将所述第一时延分解成多个时延段包括：
所述SC获得所述多个DC中每个DC所管理的域内的路径中的时延最小的路径的第二时延；
所述SC按照获得的多个所述第二时延之间的比例将所述第一时延分解成所述多个时延段。
在一个可能的设计中，当所述业务类型标识的业务为对时延不敏感的业务时，所述SC选择的所述时延分配方式为最大时延比率分配(Maximum Latency Rate Distribution，MARD)方式，所述SC根据所述MARD方式将所述第一时延分解成多个时延段包括：
所述SC获得所述多个DC中每个DC所管理的域内的路径中的时延最大的路径的第三时延；
所述SC按照获得的多个所述第三时延之间的比例将所述第一时延分解成所述多个时延段。
在一个可能的设计中，当所述业务类型标识的业务为传输文件大小超过预设值的业务时，所述SC选择的所述时延分配方式为平均光纤利用率分配(Average Fiber Utilization Rate Distribution，AFRD)方式，所述SC根据所述AFRD方式将所述第一时延分解成多个时延段包括：
所述SC获得所述多个DC中每个DC所管理的域内的光纤利用率；
所述SC按照获得所述多个光纤利用率之间的比例将所述第一时延分解成所述多个时延段。
AFRD方式能够反映域内的链路使用情况，使业务路径迭代时网络利用率更优。
在一个可能的设计中，当所述业务类型标识的业务为大数据业务时，所述SC选 择的所述时延分配方式为总流量占比比率分配(Bandwidth Occupied Rate Distribution，BORD)方式，所述SC根据所述BORD方式将所述第一时延分解成多个时延段包括：
所述SC获得所述多个DC中每个DC所管理的域内的流量，并根据获得的所述流量分别计算所述每个DC的流量占所述每个DC总带宽的比例；
所述SC按照获得多个所述每个DC的流量占所述每个DC总带宽的比例之间的比例将所述第一时延分解成所述多个时延段。
BORD方式能够反映域内的链路使用情况，使业务路径迭代时网络利用率更优。
在一个可能的设计中，所述SC选择的所述时延分配计算方式为指定时延比率分配方式，其中SC按照用户指定的比例将所述第一时延分解成所述多个时延段。
根据本发明实施例的方法，对于不同的业务类型，SC能够选择不同的时延比率分配方法，这样可以动态的分配转发路径的时延约束，满足不同的业务需求。
第二方面，本申请实施例提供了一种网络中建立转发路径的SC，所述网络包括所述SC和多个域控制器DC，所述SC管理所述多个DC，所述多个DC中的每个DC分别管理一个域，该SC包括：
接收单元，用于接收业务请求消息，所述业务请求消息包含业务类型和第一时延，所述业务请求消息请求所述SC在所述网络中建立用于传输所述业务类型对应业务的转发路径，所述转发路径的时延小于或等于所述第一时延；
选择单元，用于根据所述业务请求消息选择时延分配方式；
处理单元，用于根据所述时延分配方式将所述第一时延分解成多个时延段，所述多个时延段与所述多个DC一一对应；
发送单元，用于分别将所述多个时延段中的每个时延段发送到与所述每个时延段对应的DC，触发所述多个DC中的每个DC在其管理的域中建立转发路径段，每个所述转发路径段的时延小于或等于与所述每个DC对应的时延段；
所述处理模块，还用于将所述多个DC建立的多个转发路径段连接形成所述转发路径。
在一个可能的设计中，所述SC还包括：
确定单元，用于确定建立所述转发路径段的所述域中的网络节点的数量；
所述选择单元，用于当所述确定单元确定建立所述转发路径段的所述域中的网络节点的数量的差值在预定范围内时，选择的所述时延分配方式为ALRD方式，其中
所述第一时延按照所述转发路径段的个数平均分解成所述多个时延段。
在一个可能的设计中，所述SC还包括：
确定单元，用于确定所述业务类型标识的业务为对时延敏感的业务；
所述选择单元，用于当所述确定单元确定所述业务类型标识的业务为对时延敏感的业务时，选择的所述时延分配方式为MIRD方式，所述SC根据所述MIRD方式将所述第一时延分解成多个时延段包括：
所述SC获得所述多个DC中每个DC所管理的域内的路径中的时延最小的路径的第二时延；
所述SC按照获得的多个所述第二时延之间的比例将所述第一时延分解成所述多个时延段。
在一个可能的设计中，所述SC还包括：
确定单元，用于确定所述业务类型标识的业务为对时延不敏感的业务；
所述选择单元，用于当所述确定单元确定所述业务类型标识的业务为对时延不敏感的业务时，选择的所述时延分配方式为MARD方式，所述SC根据所述MARD方式将所述第一时延分解成多个时延段包括：
所述SC获得所述多个DC中每个DC所管理的域内的路径中的时延最大的路径的第三时延；
所述SC按照获得的多个所述第三时延之间的比例将所述第一时延分解成所述多个时延段。
在一个可能的设计中，所述SC还包括：
确定单元，用于确定所述业务类型标识的业务为传输文件大小超过预设值的业务；
所述选择单元，用于当所述确定单元确定所述业务类型标识的业务为传输文件大小超过预设值的业务时，选择的所述时延分配方式为AFRD方式，所述SC根据所述AFRD方式将所述第一时延分解成多个时延段包括：
所述SC获得所述多个DC中每个DC所管理的域内的光纤利用率；
所述SC按照获得所述多个光纤利用率之间的比例将所述第一时延分解成所述多个时延段。
在一个可能的设计中，所述SC还包括：
确定单元，用于确定所述业务类型标识的业务为大数据业务；
所述选择单元，用于当所述确定单元确定所述业务类型标识的业务为大数据业务时，选择的所述时延分配方式为BORD方式，所述SC根据所述BORD方式将所述第一时延分解成多个时延段包括：
所述SC获得所述多个DC中每个DC所管理的域内的流量，并根据获得的所述流量分别计算所述每个DC的流量占所述每个DC总带宽的比例；
所述SC按照获得多个所述每个DC的流量占所述每个DC总带宽的比例之间的比例将所述第一时延分解成所述多个时延段。
在一个可能的设计中，所述选择单元选择的所述时延分配方式为指定时延比率方式，其中所述时延按照用户指定的比例分解成所述多个时延段。
第三方面，本申请实施例提供了一种网络中建立转发路径的方法，所述网络包括SC和多个DC，所述SC管理所述多个DC，所述多个DC中的每个DC分别管理一个域，所述方法包括：
所述DC接收所述SC发送的时延段，所述时延段为所述SC根据业务请求消息选择时延分配方式后，将所述业务请求消息中的第一时延分解而成；
所述DC在管理的域中建立转发路径段，所述转发路径段的时延小于或等于与所述时延段；
所述DC将所述转发路径段的信息向所述SC发送，触发SC将所述多个DC建立的多个所述转发路径段连接形成所述转发路径。
在上述技术方案中，所建立的转发路径的时延可以根据业务需求进行定制和智能匹配，通过对业务类型进行判断，进而选择适合此业务的分配方式，解决了现有技术 中没有合适的原则来分配转发路径的时延约束的技术问题，上述方案能够使业务路径迭代时网络利用率更优，以及使得网络优化更能满足客户的需求。
在一个可能的设计中，当所述多个DC中的每个DC所管理的域内的网络节点的数量的差值在预定范围内时，所述SC选择的所述时延分配方式为ALRD方式，所述SC根据所述ALRD方式将所述第一时延分解成所述时延段包括：
所述SC按照所述DC的个数将所述第一时延平均分解成所述时延段。
在一个可能的设计中，当所述业务类型标识的业务为对时延敏感的业务时，所述SC选择的所述时延分配方式为MIRD方式，所述SC根据所述MIRD方式将所述第一时延分解成所述时延段包括：
所述SC获得所述多个DC中每个DC所管理的域内的路径中的时延最小的路径的第二时延；
所述SC按照获得的多个所述第二时延之间的比例将所述第一时延分解成所述时延段。
在一个可能的设计中，当所述业务类型标识的业务为对时延不敏感的业务时，所述SC选择的所述时延分配方式为MARD方式，所述SC根据所述MARD方式将所述第一时延分解成所述时延段包括：
所述SC获得所述多个DC中每个DC所管理的域内的路径中的时延最大的路径的第三时延；
所述SC按照获得的多个所述第三时延之间的比例将所述第一时延分解成所述时延段。
在一个可能的设计中，当所述业务类型标识的业务为传输文件大小超过预设值的业务时，所述SC选择的所述时延分配方式为AFRD方式，所述SC根据所述AFRD方式将所述第一时延分解成所述时延段包括：
所述SC获得所述多个DC中每个DC所管理的域内的光纤利用率；
所述SC按照获得所述多个光纤利用率之间的比例将所述第一时延分解成所述时延段。
AFRD方式能够反映域内的链路使用情况，使业务路径迭代时网络利用率更优。
在一个可能的设计中，当所述业务类型标识的业务为大数据业务时，所述SC选择的所述时延分配方式为BORD方式，所述SC根据所述BORD方式将所述第一时延分解成所述时延段包括：
所述SC获得所述多个DC中每个DC所管理的域内的流量，并根据获得的所述流量分别计算所述每个DC的流量占所述每个DC总带宽的比例；
所述SC按照获得多个所述每个DC的流量占所述每个DC总带宽的比例之间的比例将所述第一时延分解成所述时延段。
BORD方式能够反映域内的链路使用情况，使业务路径迭代时网络利用率更优。
在一个可能的设计中，所述SC选择的所述时延分配计算方式为指定时延比率分配方式，其中所述SC按照用户指定的比例将所述第一时延分解成所述时延段。
第四方面，本申请实施例提供了一种网络中建立转发路径的DC，所述网络包括超级控制器SC和多个所述DC，所述SC管理所述多个DC，所述多个DC中的每个DC 分别管理一个域，所述DC包括：
接收单元，用于接收所述SC发送的时延段，所述时延段为所述SC根据业务请求消息选择时延分配方式后，将所述业务请求消息中的第一时延分解而成；
建立单元，用于在所述域中建立转发路径段，所述转发路径段的时延小于或等于与所述时延段；
发送单元，用于将所述转发路径段的信息向所述SC发送，触发SC将所述建立单元建立的多个所述转发路径段连接形成所述转发路径。
第五方面，本申请实施例提供了一种建立转发路径的网络系统，包括SC、第一DC和第二DC，所述SC管理所述第一DC和所述第二DC，所述第一DC和所述第二DC分别管理一个域，
所述SC，用于接收业务请求消息，所述业务请求消息包含业务类型和第一时延，所述业务请求消息用于请求所述SC在所述网络中建立用于传输所述业务类型对应业务的转发路径，所述转发路径的时延小于或等于所述第一时延，根据所述业务请求消息选择时延分配方式，根据所述时延分配方式将所述第一时延分解成所述第一时延段和所述第二时延段，并将所述第一时延段向所述第一DC发送，将所述第二时延段向所述第二DC发送；
所述第一DC，用于接收所述SC发送的所述第一时延段，并根据所述第一时延段在其管理的域内建立第一转发路径段，所述第一转发路径段的时延小于等于所述第一时延段，将所述第一转发路径的信息发送到所述SC；
所述第二DC，用于接收所述SC发送的所述第二时延段，并根据所述第二时延段在其管理的域内建立第二转发路径段，所述第二转发路径段的时延小于等于所述第二时延段，将所述第二转发路径的信息发送到所述SC；
所述SC，还用于根据所述第一转发路径的信息和所述第二转发路径的信息将所述第一转发路径段和所述第二转发路径段连接形成所述转发路径。
第六方面，本申请实施例提供了一种网络设备，网络设备包括：处理器、网络接口和存储器。存储器可以用于存储程序代码，处理器用于调用存储器中的程序指令执行前述方面设计的方法，具体执行步骤可以参见前述的各个方面，此处不再赘述。
第七方面，本申请实施例提供了一种计算机存储介质，用于储存为上述网络设备所用的计算机软件指令，其包含用于执行上述方面所设计的程序。
附图说明
图1是本发明实施例提供的现有技术中端到端的网络结构示意图；
图2是根据本发明实施例提供的一种建立转发路径的方法的应用网络场景示意图；
图3是根据本发明实施例提供的一种中建立转发路径的方法的流程图；
图4是根据本发明实施例提供的一种建立转发路径的方法的应用场景示意图；
图5是根据本发明实施例提供的不同业务类型对应的时延分配方法的示意图；
图6A是根据本发明实施例提供的一种建立转发路径的方法的应用场景示意图；
图6B是根据本发明实施例提供的一种建立转发路径的方法的应用场景示意图；
图7是根据本发明实施例提供的一种建立转发路径的控制器的结构示意图；
图8是根据本发明实施例提供的一种中建立转发路径的方法的流程图；
图9是根据本发明实施例提供的一种建立转发路径的控制器的结构示意图；
图10是根据本发明实施例提供的一种建立转发路径的网络系统的示意图；
图11是根据本发明实施例提供的一种网络设备的结构示意图。
具体实施方式
下面将结合附图，对本发明的实施例进行详细说明。
图2为本申请实施例提供的一种网络中建立转发路径的方法的应用网络场景示意图。图2显示了由物理网络1、超级控制器2、单域控制器3、协同器4、Utraffic5组成的网络系统。其中协同器用于E2E业务下发，界面呈现；SC的功能包括：E2E业务分解，分解为单域的配置模型；网络资源管理，管理网络资源(Route-Distinguisher，RD)、RT(Route-Target)、PW-ID(Password-Identification)等)；拓扑管理，全网物理拓扑，L3拓扑，和业务拓扑状态管理；隧道管理，E2E标签交换路径(Label Switched Path，LSP)的规划管理；DC的功能包括：网元管理，物理网元发现，Inventory信息获取等；网元资源管理，管理网元资源；拓扑管理，域内物理拓扑发现，域内L3拓扑发现，状态更新等；隧道管理，域内TE隧道管理；路径计算，域内TE路径计算；而Utraffic 5用于流量监控，下发调优。
下面结合一个的例子来具体说明现有技术中存在的技术问题。
目前软件定义网络(Software Defined Network，SDN)跨域业务按需时延(Latency On Demand，LOD)尚未支持带E2E时延约束的业务发放。假设在图2所示的跨2个AS域的网络中，用户需要部署一个10ms时延要求的，从AS-1到AS-2的虚拟专用局域网业务(Virtual Private LAN Service，VPLS)专线业务，现有技术整个端到端性能指标的2/3分配给各个自治域的内部网络，整个端到端性能指标的1/3分配给连接链路。
则每个运营商网络分配的性能指标计算公式：
单个运营商网络的性能指标＝端到端性能指标*(2/3+1/3*0.35)*1/2。
网间互连链路所分配的性能指标计算公式：
网间互连链路的性能指标＝端到端性能指标*1/3*0.65。
上述的分配方法没有考虑网络的具体情况，比如域AS-1是一个比较小的网络，域AS-2是一个规模比较大的网络，按照上面的分配方法，不同AS域分配的时延性能指标都是一样的。这样的分配方法显然是不合理的，且是静态的，无法根据业务和网络情况进行在线调整。为解决上述技术问题，本申请实施例提供了一种网络中建立转发路径的方法。
为了清楚的阐述本发明的具体实施例，以包含了三个域的物理网络为例，介绍了下面SDN跨域网络支持带E2E时延约束的VPN业务发放的方法流程。图3是根据本发明第一实施例的网络中建立转发路径的方法的流程图。
如图3所示，该方法包括：
S301，SC接收业务请求消息，所述业务请求消息包含业务类型和第一时延，所述业务请求消息用于请求所述SC在所述网络中建立用于传输所述业务类型对应业务的转发路径，所述转发路径的时延小于或等于所述第一时延；
S302，SC根据所述业务请求消息选择时延分配方式；
S303，SC根据所述时延分配方式将所述第一时延分解成多个时延段，所述多个时延段与所述多个DC一一对应；
S304，SC分别将所述多个时延段中的每个时延段发送到与所述每个时延段对应的DC，触发所述多个DC中的每个DC在其管理的域中建立转发路径段，每个所述转发路径段的时延小于或等于与所述每个DC对应的时延段；
S305，SC将所述多个DC建立的多个转发路径段连接形成所述转发路径。
这里SC将多个转发路径段连接形成转发路径可以通过LSP拼接(stitching)的方式，将各个域中的LSP连接在一起形成一整条转发路径。
在一个可能的设计中，当所述多个DC中的每个DC所管理的域内的网络节点的数量的差值在预定范围内时，所述SC选择的所述时延分配方式为ALRD方式，所述SC根据所述ALRD方式将所述第一时延分解成多个时延段包括：
所述SC按照所述转发路径段的个数将所述第一时延平均分解成所述多个时延段。
在一个可能的设计中，当所述业务类型标识的业务为对时延敏感的业务时，所述SC选择的所述时延分配方式为MIRD方式，所述SC根据所述MIRD方式将所述第一时延分解成多个时延段包括：
所述SC获得所述多个DC中每个DC所管理的域内的路径中的时延最小的路径的第二时延；
所述SC按照获得的多个所述第二时延之间的比例将所述第一时延分解成所述多个时延段。
在一个可能的设计中，当所述业务类型标识的业务为对时延不敏感的业务时，所述SC选择的所述时延分配方式为MARD方式，所述SC根据所述MARD方式将所述第一时延分解成多个时延段包括：
所述SC获得所述多个DC中每个DC所管理的域内的路径中的时延最大的路径的第三时延；
所述SC按照获得的多个所述第三时延之间的比例将所述第一时延分解成所述多个时延段。
在一个可能的设计中，当所述业务类型标识的业务为传输文件大小超过预设值的业务时，所述SC选择的所述时延分配方式为AFRD方式，所述SC根据所述AFRD方式将所述第一时延分解成多个时延段包括：
所述SC获得所述多个DC中每个DC所管理的域内的光纤利用率；
所述SC按照获得所述多个光纤利用率之间的比例将所述第一时延分解成所述多个时延段。
在一个可能的设计中，当所述业务类型标识的业务为大数据业务时，所述SC选择的所述时延分配方式为BORD方式，所述SC根据所述BORD方式将所述第一时延分解成多个时延段包括：
所述SC获得所述多个DC中每个DC所管理的域内的流量，并根据获得的所述流量分别计算所述每个DC的流量占所述每个DC总带宽的比例；
所述SC按照获得多个所述每个DC的流量占所述每个DC总带宽的比例之间的比 例将所述第一时延分解成所述多个时延段。
以下，参照附图4和图5详细描述根据本发明第一实施例的建立转发路径的方法。
如图4所示，整个网络包括了SC和3个DC以及3个DC分别管理的3个AS域(SC和DC未示出)，其中各个域的信息如表1所示：
表1
域名称 AS编号 节点规模 节点时延(ms)
Domain1 As-1 Nd1 T node
Domain2 As-2 Nd2 T node
Domain3 As-3 Nd3 T node
根据上述信息，可以计算出
Figure PCTCN2018091688-appb-000001
例如：
Td1＝Node(pe1)+node(abr1)+node(asbr1)+link(pe1-abr1)+link(abr1-asbr1)
Td2＝node(pe1)+node(abr1)+node(abr2)+node(asbr2)+node(asbr1)+link(pe1-abr1)+lin k(abr1-abr2)+link(abr2-asbr2)+link(asbr2-asbr1)
表2是根据上述公式计算出的各个域链路时延属性。
表2：
Figure PCTCN2018091688-appb-000002
Figure PCTCN2018091688-appb-000003
下面以在PE1和PE3之间布放跨3个AS域的L3VPN业务，时延要求为Tsvc的情况举例说明。该L3VPN业务的下发步骤如下：
首先，北向应用程序(APP)通过报文下发带有时延约束的业务。业务报文通过restconf接口下发到SC控制器。SC控制restserver接收到北向请求之后，然后把北向请求转发到L3服务管理(Service Management，SM)模块，L3SM模块把互联网工程任务组(Internet Engineering Task Force，IETF)模型转换成稳定的NetL3VPN模型。
SC上的NetL3vpn对北向报文进行解释，申请网络级资源RD、RT，如果北向报文里面携带的“latency”，则提取业务特征后，根据业务特征调用POLICY模块获取时延分配方法。
这里，业务特征的提取，方式可以有如下两种：
第一种是SC控制器根据北向报文中的入参path可以提取业务类型。例如“path＝URL：http://{{url}}/restconf/config/ietf-l3vpn-svc:l3vpn-svc/vpn-services”，由此可以识别业务类型为L3VPN类型。
另一种可以通过网络用户接口(Website User Interface，WEBUI)界面预先配置，当业务下发时，北向APP可以指定所要创建的业务类型为指定的业务类型。
根据本发明实施例的网络中建立转发路径的方法，在识别了业务类型后，需要根据不同的业务类别选择对应的分配方法。图5是根据本发明实施例的不同的业务类型所对应的不同的分配方法的示意图。
各种分配方法的介绍及计算公式如下：
1)平均时延比率分配(Average Latency Rate Distribution，简称：ALRD)
Figure PCTCN2018091688-appb-000004
其中，Ti为第i个域分配的时延，TTotal为隧道的总时延约束，DTotal为隧道的自治域总和。
如用户需要在一个跨3个AS域内创建一个时延约束为10ms的专线，按照平均时延，则每个AS域的时延约束为3.3ms。
这种分配方式适用于隧道跨的自治域的网络规模相当的场景，比如网络节点的数量差异在预定范围，如10％以内的情况。如一个核心路由器下挂两个规模相当的城域网，这两个城域网之间需要创建一个时延为10ms的隧道时，这种分配方式可用于例如互联网协议安全性(Internet Protocol Security，IPSEC)等业务。
2)按照用户指定比率分配(Customized Latency Rate Distribution简称：CLRD)
按照用户指定的时延比率分配。如用户指定2:3:5，则总时延为10ms的隧道，各个域分配的时延分别为2ms，3ms，5ms。
这种分配方式一般是网络管理员结合自己的网络规划、时延规划，同时根据网络负载监控的情况，实时在线时延分配方法的场景。如网络管理员根据网络性能监测结果，发现某个域负载很大，某域负载很小，这时就可以手工调整时延分配比例，定制化时延分配。
3)按各AS域最小时延比率分配(Minimum Latency Rate Distribution简称：MIRD)
T Min＝Min(Delay j)
其中，TMin是域内所有的路径中时延最小的路径的时延值，Delayj是第j条路径的时延。
如各AS域的最小时延比为1:2:2，则总时延为10ms的隧道，各个域分配的时延分别为2ms，4ms，4ms。
这个分配方法，时延小的域分配迭代时延少，时延大的分配迭代时延大，兼顾轻载网络和重载网络时延，这种方法是一种较为简单的分配方式且能达到网络负载得到调优。这种分配适用于对时延约束比较敏感的业务，例如L3VPN，L2VPN和EVPN等。
在一种可能的实现方式中，所述对时延敏感的业务具体是指传输所述业务的时延不能大于特定阈值，该特定阈值可以有用户根据实际情况指定。
4)按照各AS域最大时延比率分配(Maximum Latency Rate Distribution简称：MARD)
T MAX＝Max(Delay j)
其中，TMAX是域内所有的路径中时延最大的路径的时延值，Delayj是第j条路径的时延。
如各AS域的最小时延比为15:10:15，则各个域分配的时延为4ms，2ms，4ms。
这种方法是比较有弹性的分配方式，最大迟延大的域分配的迭代时延多，这样域内可选的路径就比较多。适用于对时延敏感度相对不是很高的业务，例如INTERNET、FTP、TELNET、EMAIL或P2P等。
5)按照域内平均光纤利用率比率分配(Average Fiber Utilization Rate Distribution简称：AFRD)
Figure PCTCN2018091688-appb-000005
其中，m是域内光纤总的条数，FURi(fiber utilization rate)是第i个域的平均光纤利用率。
如果各个域的平均光纤利用率分别为20％：15％：15％，则各个域分配的时延为4ms，3ms，3ms。平均光纤利用率的计算方法：(光纤1通过流量/光纤1最大带宽+…+光纤n通过流量/光纤n最大带宽)/n*100％。
这种分配方式从域内光纤利用比率来分配，有利于提升域内的带宽利用率，可用于传输大文件的跨域业务，所谓的大文件是大小超过预设值的文件。例如视频(VIDEO)和网络电话(Voice over Internet Protocol，VOIP)等用于传输图像和/或声音的跨域业务。
6)按照域内总流量占比比率(Bandwidth Occupied Rate Distribution简称：BORD)
Figure PCTCN2018091688-appb-000006
其中，m为域内总的链接数，BORi(bandwidth occupied rate)是第i个域的总流量占比，Flowj为第j条链路通过的流量，BandWidthj是第j条链路的最大物理带宽。
如果各个域内，总流量占比分别为10％：30％：60％，则各个域分配的时延为1ms，3ms，6ms。域内总流量占比的计算方法：(链路1通过的流量+…+链路n通过的流量)/(链路1的最大带宽+…+链路n的最大带宽)*100％。
这种分配方式把域看做一个黑匣子，通过的总流量比上总的容量比率，类似于“域内平均光纤利用率”，都是从网络带宽利用率角度来考虑分配方法。因此，在用于普通数据业务，特别是大数据(Big Data)的跨域业务时，如VXLAN，这种分配方式比“域内平均光纤利用率”相比更好一些。
平均光纤利用率、总流量占比方法，能反映域内的链路使用情况，这两种方法能够使业务路径迭代时网络利用率更优。
图5是根据本发明的实施例的时延分配方法和具体业务类型的对应关系图。由图5可知，针对本实施例的L3VPN业务，应当选择“最小时延比率”的分配方法。根据上述的最小时延比率的分配方法的计算式，可以得到在第一个域内，T min＝Td1ms，在第二个域内，T min＝Td9ms，第三个域内，T min＝Td17ms。
从而，可以计算得到各个域内的迭代时延约束。在本实施例中，所要求的时延Tsvc将按照Td1：Td9：Td17的比率来分配。因此，各个域分配的时延分别为Tsvc*Td1/(Td1+Td9+Td17)，Tsvc*Td9/(Td1+Td9+Td17)以及Tsvc*Td17/(Td1+Td9+Td17)。
在根据选择的分配方法对各个域的时延约束进行分配的步骤之后，进行通信链路的建立。具体的，SC上的NetL3vpn对模型业务基于DC粒度进行分割，分解成域级的网络数据并下发给DC控制器。
相应地，DC接收来自SC的网络数据，调用SNC组件，基于时延要求约束动态创建TE隧道作为L3VPN的承载隧道，即创建时延小于等于所分配的时延段的转发路径。对于第一个域，按照Tsvc*Td1/(Td1+Td9+Td17)ms的时延约束去动态创建一条TE(Traffic Engineering，流量工程)隧道；对于第二个域，按照Tsvc*Td9/(Td1+Td9+Td17)ms的时延约束去动态创建一条TE隧道；对于第三个域，按 照Tsvc*Td17/(Td1+Td9+Td17)ms的时延约束去动态创建一条TE隧道。
完成TE隧道的创建之后，DC将相关的TE隧道的信息向SC发送，触发SC执行LSP拼接，从而将各个域中的TE隧道连接形成一整条转发路径。
在本实施例中，虽然迭代隧道时时延约束分别为Tsvc*Td1/(Td1+Td9+Td17)ms,Tsvc*Td9/(Td1+Td9+Td17)ms,Tsvc*Td17/(Td1+Td9+Td17)ms，实际迭代到的隧道时延遵循最小时延优选原则。注意，这里迭代的隧道为双向共路隧道。
最终，三个域迭代的隧道如图4中粗实线所示。
下面，对照图6A，以实际参数为例，对根据本发明第一实施例的方法进行描述。
图6A显示了与图4相同的物理网络，包含有3个AS域。其中各个域的信息如表3所示：
表3
Figure PCTCN2018091688-appb-000007
并且根据图6A中标示的时延值可以计算出如表4所示的各个域的链路时延：
表4
Figure PCTCN2018091688-appb-000008
Figure PCTCN2018091688-appb-000009
同样以在PE1和PE3之间布放跨3个AS域的L3VPN业务为例，时延要求为20ms。如上所述，针对本实施例的L3VPN业务，利用最小时延比率分配方法，计算出各个域中的最小时延分别是2.45ms，2.55ms和0.9ms。那么各个域按照2.45：2.55：0.9比率分配时延，各个域分配的时延分别为8.3ms，8.7ms和3ms。然后，SC将这三个时延段分别发送给对应的DC，从而触发DC在其管理的域内建立转发路径，即隧道。
对于第一个域，按照8.3ms的时延约束去动态创建一条TE隧道；对于第二个域，按照8.7ms的时延约束去动态创建一条TE隧道；对于第三个域，按照3ms的时延约束去动态创建一条TE隧道。
隧道约束带有时延约束，则隧道创建方式采用精确匹配方式。即隧道迭代管理模块按照时延约束条件选择一条精确匹配时延要求的路径。如果没有找到，则新建一条满足时延延期的隧道。如要创建一条时延约束为50ms的隧道，PCE模块根据约束条件去算路，路径的时延为路径上节点和链路时延的线性加和。如果满足时延要求的路径有多条，则可以选择时延累加和最小的一条。
最终，三个域迭代的隧道如图6B中的粗实线61、62和63所示。由于迭代隧道的时延约束分别是8.3ms，8.7ms和3ms，各个DC创建的转发路径的时延需要小于等于各自所分配的时延约束段。实际使用中，迭代到的隧道时延可遵循最小时延优选原则，最终创建的隧道61的时延是2.45ms，隧道62的时延是2.55ms，隧道63的时延是0.9ms。
根据上述实施例，本发明还提供了另一种网络中建立转发路径的方法，其中该网络包括如上所述的SC和多个DC。
其中，该方法包括，DC接收SC发送的时延段，所述时延段为所述SC根据业务请求消息选择时延分配方式后，将所述业务请求消息中的第一时延分解而成；
DC在所述域中建立转发路径段，所述转发路径段的时延小于或等于与所述时延段；
DC将所述转发路径段的信息向所述SC发送，触发SC将多个所述转发路径段连 接形成所述转发路径。
进一步的，根据本发明第一实施例的方法还可以通过对网络进行监控，从而判断各个域的时延分配是否满足了业务需求。
如果判断为否，引入用户的输入，对时延参数进行修改，然后对各个域的时延约束进行重新分配。如果判断为是，则维持当前的各个域的时延约束。
根据本发明实施例提供的方法支持用户在线修改功能，当业务时延不满足业务需求时，可以引入用户的输入，从而在线修改时延参数值。时延参数更新后，控制器域内将根据新时延进行迭代满足时延约束的隧道，迭代成功之后，VPN就切换到新的路径上。
如图2所示，用户通过Utraffic 5监控网络链路利用率、时延、抖动、丢包率等性能指标，对比预期来调整时延分配方法。DC控制器通过Utraffic 5提供的接口获取单域平均光纤利用率、单域总流量占比、链路时延、节点时延等实时网络性能数据。因此，用户通过Utraffic 5监测E2E业务时延指标，当发现不满足业务时，可以触发控制器动态在线修改链路的时延，触发业务路径的重现迭代，重新识别并分配各个域的时延分配，使得业务流量路径调整到更优的路径。另外，在监测后发现业务时延指标不满足业务时，作为另一种选择，可以引入用户的在线输入。用户可以在线修改业务时延参数，通过引入人工定制的时延约束修改，使得网络优化更能满足客户的需求。
现有的技术节点、链路时延属性都是静态配置。预设的时延方法实时性强。最小时延、最大时延，光纤利用率、节点时延、链路迟延等网络性能数据可通过监控器(如Utraffic)提供的接口获取并实时或周期刷新控制器节点和链路属性，因此具备实时性，能反映了网络某时段的真实状态。
参见图7，本申请实施例还提供了一种网络中建立转发路径的SC700，如图7所示，该SC包括：
接收单元701，用于接收业务请求消息，所述业务请求消息包含业务类型和第一时延，所述业务请求消息请求所述SC在所述网络中建立用于传输所述业务类型对应业务的转发路径，所述转发路径的时延小于或等于所述第一时延；
选择单元702，用于根据所述业务请求消息选择时延分配方式；
处理单元703，用于根据所述时延分配方式将所述第一时延分解成多个时延段，所述多个时延段与所述多个DC一一对应；
发送单元704，用于分别将所述多个时延段中的每个时延段发送到与所述每个时延段对应的DC，触发所述多个DC中的每个DC在其管理的域中建立转发路径段，每个所述转发路径段的时延小于或等于与所述每个DC对应的时延段；
所述处理单元703，还用于将所述多个DC建立的多个转发路径段连接形成所述转发路径。
SC700所实施的动作以及实现的功能与前述实施例所描述的建立转发路径的方法一致，因此不再赘述。
根据本申请实施例的SC700还包括：
确定单元，用于确定建立所述转发路径段的所述域中的网络节点的数量；
所述选择单元，用于当所述确定单元确定建立所述转发路径段的所述域中的网络 节点的数量的差值在预定范围内时，选择的所述时延分配方式为ALRD方式；
所述处理单元根据所述时延分配方式将所述第一时延分解成多个时延段包括：
所述处理单元将所述第一时延按照所述多个时延段的个数平均分解成所述多个时延段。
根据本申请实施例的SC700还包括：
确定单元，用于确定所述业务类型标识的业务是否为对时延敏感的业务；
所述选择单元，用于当所述确定单元确定所述业务类型标识的业务为对时延敏感的业务时，选择的所述时延分配方式为MIRD方式；
所述处理单元根据所述MIRD方式将所述第一时延分解成多个时延段包括：
所述处理单元获得所述多个DC中每个DC所管理的域内的路径中的时延最小的路径的第二时延；
所述处理单元按照获得的多个所述第二时延之间的比例将所述第一时延分解成所述多个时延段。
根据本申请实施例的SC700还包括：
确定单元，用于确定所述业务类型标识的业务是否为对时延不敏感的业务；
所述选择单元，用于当所述确定单元确定所述业务类型标识的业务为对时延不敏感的业务时，选择的所述时延分配方式为MARD方式；
所述处理单元根据所述MARD方式将所述第一时延分解成多个时延段包括：
所述处理单元获得所述多个DC中每个DC所管理的域内的路径中的时延最大的路径的第三时延；
所述处理单元按照获得的多个所述第三时延之间的比例将所述第一时延分解成所述多个时延段。
根据本申请实施例的SC700还包括：
确定单元，用于确定所述业务类型标识的业务是否为传输文件大小超过预设值的业务；
所述选择单元，用于当所述确定单元确定所述业务类型标识的业务为传输文件大小超过预设值的业务时，选择的所述时延分配方式为AFRD方式；
所述处理单元根据所述AFRD方式将所述第一时延分解成多个时延段包括：
所述处理单元获得所述多个DC中每个DC所管理的域内的光纤利用率；
所述处理单元按照获得所述多个光纤利用率之间的比例将所述第一时延分解成所述多个时延段。
根据本申请实施例的SC700还包括：
确定单元，用于确定所述业务类型标识的业务是否为大数据业务；
所述选择单元，用于当所述确定单元确定所述业务类型标识的业务为大数据业务时，选择的所述时延分配方式为BORD方式；
所述处理单元根据所述BORD方式将所述第一时延分解成多个时延段包括：
所述处理单元获得所述多个DC中每个DC所管理的域内的流量，并根据获得的所述流量分别计算所述每个DC的流量占所述每个DC总带宽的比例；
所述处理单元按照获得多个所述每个DC的流量占所述每个DC总带宽的比例之 间的比例将所述第一时延分解成所述多个时延段。
根据本申请实施例的SC700，其中所述选择单元选择的所述时延分配方式为指定时延比率方式，其中所述处理单元按照用户指定的比例将所述第一时延分解成所述多个时延段。
根据本发明实施例，还提供了另一种网络中建立转发路径的方法，所述网络包括超级控制器SC和多个域控制器DC，所述SC管理所述多个DC，所述多个DC中的每个DC分别管理一个域。如图8所示，该方法包括：
S801，所述DC接收所述SC发送的时延段，所述时延段为所述SC根据业务请求消息选择时延分配方式后，将所述业务请求消息中的第一时延分解而成；
S802，所述DC在管理的域中建立转发路径段，所述转发路径段的时延小于或等于与所述时延段；
S803，所述DC将所述转发路径段的信息向所述SC发送，触发SC将所述多个DC建立的多个所述转发路径段连接形成所述转发路径。
在一个可能的设计中，当所述多个DC中的每个DC所管理的域内的网络节点的数量的差值在预定范围内时，所述SC选择的所述时延分配方式为ALRD方式，所述SC根据所述ALRD方式将所述第一时延分解成所述时延段包括：
所述SC按照所述DC的个数将所述第一时延平均分解成所述时延段。
在一个可能的设计中，当所述业务类型标识的业务为对时延敏感的业务时，所述SC选择的所述时延分配方式为MIRD方式，所述SC根据所述MIRD方式将所述第一时延分解成所述时延段包括：
所述SC获得所述多个DC中每个DC所管理的域内的路径中的时延最小的路径的第二时延；
所述SC按照获得的多个所述第二时延之间的比例将所述第一时延分解成所述时延段。
在一个可能的设计中，当所述业务类型标识的业务为对时延不敏感的业务时，所述SC选择的所述时延分配方式为MARD方式，所述SC根据所述MARD方式将所述第一时延分解成所述时延段包括：
所述SC获得所述多个DC中每个DC所管理的域内的路径中的时延最大的路径的第三时延；
所述SC按照获得的多个所述第三时延之间的比例将所述第一时延分解成所述时延段。
在一个可能的设计中，当所述业务类型标识的业务为传输文件大小超过预设值的业务时，所述SC选择的所述时延分配方式为AFRD方式，所述SC根据所述AFRD方式将所述第一时延分解成所述时延段包括：
所述SC获得所述多个DC中每个DC所管理的域内的光纤利用率；
所述SC按照获得所述多个光纤利用率之间的比例将所述第一时延分解成所述时延段。
在一个可能的设计中，当所述业务类型标识的业务为大数据业务时，所述SC选择的所述时延分配方式为BORD方式，所述SC根据所述BORD方式将所述第一时延 分解成所述时延段包括：
所述SC获得所述多个DC中每个DC所管理的域内的流量，并根据获得的所述流量分别计算所述每个DC的流量占所述每个DC总带宽的比例；
所述SC按照获得多个所述每个DC的流量占所述每个DC总带宽的比例之间的比例将所述第一时延分解成所述时延段。
在一个可能的设计中，所述SC选择的所述时延分配计算方式为指定时延比率分配方式，其中所述SC按照用户指定的比例将所述第一时延分解成所述时延段。
进一步的，本申请的实施例还提供了一种网络中建立转发路径的域控制器DC，所述网络包括超级控制器SC和多个所述DC，所述SC管理所述多个DC，所述多个DC中的每个DC分别管理一个域。如图9所示，所述DC900包括：
接收单元901，用于接收所述SC发送的时延段，所述时延段为所述SC根据业务请求消息选择时延分配方式后，将所述业务请求消息中的第一时延分解而成；
建立单元902，用于在所述域中建立转发路径段，所述转发路径段的时延小于或等于与所述时延段；
发送单元903，用于将所述转发路径段的信息向所述SC发送，触发SC将所述建立单元建立的多个所述转发路径段连接形成所述转发路径。
DC900所实施的动作以及实现的功能与前述实施例所描述的建立转发路径的方法一致，因此不再赘述。
根据本发明实施例，还提供了一种建立转发路径的网络系统1000。
如图10所述，该系统1000包括SC1001、第一DC1002和第二DC1003，SC1001管理第一DC1002和第二DC1003，第一DC1002和第二DC1003分别管理一个域，其中，
SC1001，用于接收业务请求消息，所述业务请求消息包含业务类型和第一时延，所述业务请求消息用于请求SC1001在所述网络中建立用于传输所述业务类型对应业务的转发路径，所述转发路径的时延小于或等于所述第一时延，根据所述业务请求消息选择时延分配方式，根据所述时延分配方式将所述第一时延分解成所述第一时延段和所述第二时延段，并将所述第一时延段向第一DC1002发送，将所述第二时延段向第二DC1003；
第一DC1002，用于接收SC1001发送的所述第一时延段，并根据所述第一时延段在其管理的域内建立第一转发路径段，所述第一转发路径段的时延小于等于所述第一时延段，将所述第一转发路径的信息发送到SC1001；
第二DC1003，用于接收SC1001发送的所述第二时延段，并根据所述第二时延段在其管理的域内建立第二转发路径段，所述第二转发路径段的时延小于等于所述第二时延段，将所述第二转发路径的信息发送到SC1001；
SC1001，还用于根据所述第一转发路径的信息和所述第二转发路径的信息将所述第一转发路径段和所述第二转发路径段连接形成所述转发路径。
该系统1000中的SC1001，第一DC1002和第二DC1003所实施的动作以及实现的功能与前述实施例所描述的建立转发路径的方法一致，因此不再赘述。
参见图11，本申请实施例还提供了一种网络设备，如图11所示，网络设备1100 可以执行前述实施例所示的方法。
网络设备1100包括：处理器1101、网络接口1102和存储器1103。处理器1101包括但不限于中央处理机(Central Processing Unit，CPU)、网络处理器(Network Processor，NP)、专用集成电路(Application-Specific Integrated Circuit，ASIC)或者可编程逻辑器件(Programmable Logic Device，PLD)中的一个或多个。上述PLD可以是复杂可编程逻辑器件(Complex Programmable Logic Device，CPLD)、现场可编程逻辑门阵列(Field-Programmable Gate Array，FPGA)或通用阵列逻辑(Generic Array Logic,GAL)。
存储器1103包括但不限于内容寻址存储器(Content-Addressable Memory，CAM)、例如三态内容寻址存储器(Ternary CAM，TCAM)或随机存取存储器(Random-Access Memory，RAM)。
所述网络接口1102可以包含普通物理接口，所述物理接口可以为Ethernet接口或异步传输模式(Asynchronous Transfer Mode,ATM)接口。处理器1101、网络接口1102和存储器1103可以集成为一个或多个独立的电路或硬件，如：ASIC。
存储器1103可以用于存储该网络设备1100的程序代码和数据，处理器1101用于调用存储器1103中的程序指令执行图3所示实施例中的方法，具体执行步骤可以参见前述实施例，此处不再赘述。
本发明的第四实施例，提供了一种计算机存储介质，用于储存为上述网络设备所用的计算机软件指令，其包含用于执行上述建立转发路径的方法所设计的程序。本领域技术人员应该可以意识到，计算机可读介质包括计算机存储介质和通信介质，其中通信介质包括便于从一个地方向另一个地方传送计算机程序的任何介质。存储介质可以是通用或专用计算机能够存取的任何可用介质，例如,存储介质可以是下述介质中的至少一种：只读存储器(Read-Only Memory，ROM)、RAM、磁碟或者光盘等各种可以存储程序代码的介质。
根据本发明的实施例提供的方法和控制器，使得时延支持可定制和智能匹配，解决了无法针对不同的业务来分配转发路径的时延约束的技术问题。虽然以上的实施例都是基于跨越的VPN业务来描述，实际上，对于其他类型的跨域业务，同样能够使用本发明所提出的方法和控制器。
本发明实施例中提到的第一时延中的“第一”只是用来做名字标识，并不代表顺序上的第一。该规则同样适用于“第二”和“第三”。
本领域技术人员应该可以意识到，在上述一个或多个示例中，本发明所描述的功能可以用硬件、软件、固件或它们的任意组合来实现。当使用软件实现时，可以将这些功能存储在计算机可读介质中或者作为计算机可读介质上的一个或多个指令或代码进行传输。计算机可读介质包括计算机存储介质和通信介质，其中通信介质包括便于从一个地方向另一个地方传送计算机程序的任何介质。存储介质可以是通用或专用计算机能够存取的任何可用介质，例如,存储介质可以是下述介质中的至少一种：只读存储器(Read-Only Memory，ROM)、RAM、磁碟或者光盘等各种可以存储程序代码的介质。
最后应说明的是：以上实施例仅用以示例性说明本申请的技术方案，而非对其限 制；尽管参照前述实施例对本申请及本申请带来的有益效果进行了详细的说明，本领域的普通技术人员应当理解：其依然可以对前述各实施例所记载的技术方案进行修改，或者对其中部分技术特征进行等同替换。

Claims (23)

  1. 一种网络中建立转发路径的方法，其特征在于，所述网络包括超级控制器SC和多个域控制器DC，所述SC管理所述多个DC，所述多个DC中的每个DC分别管理一个域，所述方法包括：
    所述SC接收业务请求消息，所述业务请求消息包含业务类型和第一时延，所述业务请求消息用于请求所述SC在所述网络中建立用于传输所述业务类型对应业务的转发路径，所述转发路径的时延小于或等于所述第一时延；
    所述SC根据所述业务请求消息选择时延分配方式；
    所述SC根据所述时延分配方式将所述第一时延分解成多个时延段，所述多个时延段与所述多个DC一一对应；
    所述SC分别将所述多个时延段中的每个时延段发送到与所述每个时延段对应的DC，触发所述多个DC中的每个DC在其管理的域中建立转发路径段，每个所述转发路径段的时延小于或等于与所述每个DC对应的时延段；
    所述SC将所述多个DC建立的多个所述转发路径段连接形成所述转发路径。
  2. 根据权利要求1所述的方法，其特征在于，当所述多个DC中的每个DC所管理的域内的网络节点的数量的差值在预定范围内时，所述SC选择的所述时延分配方式为平均时延分配ALRD方式，所述SC根据所述ALRD方式将所述第一时延分解成多个时延段包括：
    所述SC按照所述DC的个数将所述第一时延平均分解成所述多个时延段。
  3. 根据权利要求1所述的方法，其特征在于，当所述业务类型标识的业务为对时延敏感的业务时，所述SC选择的所述时延分配方式为最小时延比率分配MIRD方式，所述SC根据所述MIRD方式将所述第一时延分解成多个时延段包括：
    所述SC获得所述多个DC中每个DC所管理的域内的路径中的时延最小的路径的第二时延；
    所述SC按照获得的多个所述第二时延之间的比例将所述第一时延分解成所述多个时延段。
  4. 根据权利要求1所述的方法，其特征在于，当所述业务类型标识的业务为对时延不敏感的业务时，所述SC选择的所述时延分配方式为最大时延比率分配MARD方式，所述SC根据所述MARD方式将所述第一时延分解成多个时延段包括：
    所述SC获得所述多个DC中每个DC所管理的域内的路径中的时延最大的路径的第三时延；
    所述SC按照获得的多个所述第三时延之间的比例将所述第一时延分解成所述多个时延段。
  5. 根据权利要求1所述的方法，其特征在于，当所述业务类型标识的业务为传输文件大小超过预设值的业务时，所述SC选择的所述时延分配方式为平均光纤利用率 分配AFRD方式，所述SC根据所述AFRD方式将所述第一时延分解成多个时延段包括：
    所述SC获得所述多个DC中每个DC所管理的域内的光纤利用率；
    所述SC按照获得所述多个光纤利用率之间的比例将所述第一时延分解成所述多个时延段。
  6. 根据权利要求1所述的方法，其特征在于，当所述业务类型标识的业务为大数据业务时，所述SC选择的所述时延分配方式为总流量占比比率分配BORD方式，所述SC根据所述BORD方式将所述第一时延分解成多个时延段包括：
    所述SC获得所述多个DC中每个DC所管理的域内的流量，并根据获得的所述流量分别计算所述每个DC的流量占所述每个DC总带宽的比例；
    所述SC按照获得多个所述每个DC的流量占所述每个DC总带宽的比例之间的比例将所述第一时延分解成所述多个时延段。
  7. 根据权利要求1所述的方法，其特征在于，所述SC选择的所述时延分配计算方式为指定时延比率分配方式，所述SC按照用户指定的比例将所述第一时延分解成所述多个时延段。
  8. 一种网络中建立转发路径的超级控制器SC，其特征在于，所述网络包括所述SC和多个域控制器DC，所述SC管理所述多个DC，所述多个DC中的每个DC分别管理一个域，所述SC包括：
    接收单元，用于接收业务请求消息，所述业务请求消息包含业务类型和第一时延，所述业务请求消息用于请求所述SC在所述网络中建立用于传输所述业务类型对应业务的转发路径，所述转发路径的时延小于或等于所述第一时延；
    选择单元，用于根据所述业务请求消息选择时延分配方式；
    处理单元，用于根据所述时延分配方式将所述第一时延分解成多个时延段，所述多个时延段与所述多个DC一一对应；
    发送单元，用于分别将所述多个时延段中的每个时延段发送到与所述每个时延段对应的DC，触发所述多个DC中的每个DC在其管理的域中建立转发路径段，每个所述转发路径段的时延小于或等于与所述每个DC对应的时延段；
    所述处理单元，还用于将所述多个DC建立的多个所述转发路径段连接形成所述转发路径。
  9. 根据权利要求8所述的SC，其特征在于，所述SC还包括：
    确定单元，用于确定建立所述转发路径段的所述域中的网络节点的数量；
    所述选择单元，用于当所述确定单元确定建立所述转发路径段的所述域中的网络节点的数量的差值在预定范围内时，选择的所述时延分配方式为平均时延分配ALRD方式；
    所述处理单元根据所述时延分配方式将所述第一时延分解成多个时延段包括：
    所述处理单元将所述第一时延按照所述多个时延段的个数平均分解成所述多个时延段。
  10. 根据权利要求8所述的SC，其特征在于，所述SC还包括：
    确定单元，用于确定所述业务类型标识的业务是否为对时延敏感的业务；
    所述选择单元，用于当所述确定单元确定所述业务类型标识的业务为对时延敏感的业务时，选择的所述时延分配方式为最小时延比率分配MIRD方式；
    所述处理单元根据所述MIRD方式将所述第一时延分解成多个时延段包括：
    所述处理单元获得所述多个DC中每个DC所管理的域内的路径中的时延最小的路径的第二时延；
    所述处理单元按照获得的多个所述第二时延之间的比例将所述第一时延分解成所述多个时延段。
  11. 根据权利要求8所述的SC，其特征在于，所述SC还包括：
    确定单元，用于确定所述业务类型标识的业务是否为对时延不敏感的业务；
    所述选择单元，用于当所述确定单元确定所述业务类型标识的业务为对时延不敏感的业务时，选择的所述时延分配方式为最大时延比率分配MARD方式；
    所述处理单元根据所述MARD方式将所述第一时延分解成多个时延段包括：
    所述处理单元获得所述多个DC中每个DC所管理的域内的路径中的时延最大的路径的第三时延；
    所述处理单元按照获得的多个所述第三时延之间的比例将所述第一时延分解成所述多个时延段。
  12. 根据权利要求8所述的SC，其特征在于，所述SC还包括：
    确定单元，用于确定所述业务类型标识的业务是否为传输文件大小超过预设值的业务；
    所述选择单元，用于当所述确定单元确定所述业务类型标识的业务为传输文件大小超过预设值的业务时，选择的所述时延分配方式为平均光纤利用率分配AFRD方式；
    所述处理单元根据所述AFRD方式将所述第一时延分解成多个时延段包括：
    所述处理单元获得所述多个DC中每个DC所管理的域内的光纤利用率；
    所述处理单元按照获得所述多个光纤利用率之间的比例将所述第一时延分解成所述多个时延段。
  13. 根据权利要求8所述的SC，其特征在于，所述SC还包括：
    确定单元，用于确定所述业务类型标识的业务是否为大数据业务；
    所述选择单元，用于当所述确定单元确定所述业务类型标识的业务为大数据业务时，选择的所述时延分配方式为总流量占比比率分配BORD方式；
    所述处理单元根据所述BORD方式将所述第一时延分解成多个时延段包括：
    所述处理单元获得所述多个DC中每个DC所管理的域内的流量，并根据获得的 所述流量分别计算所述每个DC的流量占所述每个DC总带宽的比例；
    所述处理单元按照获得多个所述每个DC的流量占所述每个DC总带宽的比例之间的比例将所述第一时延分解成所述多个时延段。
  14. 根据权利要求8所述的SC，其特征在于，所述选择单元选择的所述时延分配方式为指定时延比率方式，其中所述处理单元按照用户指定的比例将所述第一时延分解成所述多个时延段。
  15. 一种网络中建立转发路径的方法，其特征在于，所述网络包括超级控制器SC和多个域控制器DC，所述SC管理所述多个DC，所述多个DC中的每个DC分别管理一个域，所述方法包括：
    所述DC接收所述SC发送的时延段，所述时延段为所述SC根据业务请求消息选择时延分配方式后，将所述业务请求消息中的第一时延分解而成；
    所述DC在管理的域中建立转发路径段，所述转发路径段的时延小于或等于与所述时延段；
    所述DC将所述转发路径段的信息向所述SC发送，触发SC将所述多个DC建立的多个所述转发路径段连接形成所述转发路径。
  16. 根据权利要求15所述的方法，其特征在于，当所述多个DC中的每个DC所管理的域内的网络节点的数量的差值在预定范围内时，所述SC选择的所述时延分配方式为平均时延分配ALRD方式，所述SC根据所述ALRD方式将所述第一时延分解成所述时延段包括：
    所述SC按照所述DC的个数将所述第一时延平均分解成所述时延段。
  17. 根据权利要求15所述的方法，其特征在于，当所述业务类型标识的业务为对时延敏感的业务时，所述SC选择的所述时延分配方式为最小时延比率分配MIRD方式，所述SC根据所述MIRD方式将所述第一时延分解成所述时延段包括：
    所述SC获得所述多个DC中每个DC所管理的域内的路径中的时延最小的路径的第二时延；
    所述SC按照获得的多个所述第二时延之间的比例将所述第一时延分解成所述时延段。
  18. 根据权利要求15所述的方法，其特征在于，当所述业务类型标识的业务为对时延不敏感的业务时，所述SC选择的所述时延分配方式为最大时延比率分配MARD方式，所述SC根据所述MARD方式将所述第一时延分解成所述时延段包括：
    所述SC获得所述多个DC中每个DC所管理的域内的路径中的时延最大的路径的第三时延；
    所述SC按照获得的多个所述第三时延之间的比例将所述第一时延分解成所述时延段。
  19. 根据权利要求15所述的方法，其特征在于，当所述业务类型标识的业务为传输文件大小超过预设值的业务时，所述SC选择的所述时延分配方式为平均光纤利用率分配AFRD方式，所述SC根据所述AFRD方式将所述第一时延分解成所述时延段包括：
    所述SC获得所述多个DC中每个DC所管理的域内的光纤利用率；
    所述SC按照获得所述多个光纤利用率之间的比例将所述第一时延分解成所述时延段。
  20. 根据权利要求15所述的方法，其特征在于，当所述业务类型标识的业务为大数据业务时，所述SC选择的所述时延分配方式为总流量占比比率分配BORD方式，所述SC根据所述BORD方式将所述第一时延分解成所述时延段包括：
    所述SC获得所述多个DC中每个DC所管理的域内的流量，并根据获得的所述流量分别计算所述每个DC的流量占所述每个DC总带宽的比例；
    所述SC按照获得多个所述每个DC的流量占所述每个DC总带宽的比例之间的比例将所述第一时延分解成所述时延段。
  21. 根据权利要求15所述的方法，其特征在于，所述SC选择的所述时延分配计算方式为指定时延比率分配方式，其中所述SC按照用户指定的比例将所述第一时延分解成所述时延段。
  22. 一种网络中建立转发路径的域控制器DC，其特征在于，所述网络包括超级控制器SC和多个所述DC，所述SC管理所述多个DC，所述多个DC中的每个DC分别管理一个域，所述DC包括：
    接收单元，用于接收所述SC发送的时延段，所述时延段为所述SC根据业务请求消息选择时延分配方式后，将所述业务请求消息中的第一时延分解而成；
    建立单元，用于建立转发路径段，所述转发路径段的时延小于或等于与所述时延段；
    发送单元，用于将所述转发路径段的信息向所述SC发送，触发SC将所述建立单元建立的多个所述转发路径段连接形成所述转发路径。
  23. 一种建立转发路径的网络系统，其特征在于，包括SC、第一DC和第二DC，所述SC管理所述第一DC和所述第二DC，所述第一DC和所述第二DC分别管理一个域，
    所述SC，用于接收业务请求消息，所述业务请求消息包含业务类型和第一时延，所述业务请求消息用于请求所述SC在所述网络中建立用于传输所述业务类型对应业务的转发路径，所述转发路径的时延小于或等于所述第一时延，根据所述业务请求消息选择时延分配方式，根据所述时延分配方式将所述第一时延分解成第一时延段和第二时延段，并将所述第一时延段向所述第一DC发送，将所述第二时延段向所述第二 DC发送；
    所述第一DC，用于接收所述SC发送的所述第一时延段，并根据所述第一时延段在其管理的域内建立第一转发路径段，所述第一转发路径段的时延小于等于所述第一时延段，将所述第一转发路径的信息发送到所述SC；
    所述第二DC，用于接收所述SC发送的所述第二时延段，并根据所述第二时延段在其管理的域内建立第二转发路径段，所述第二转发路径段的时延小于等于所述第二时延段，将所述第二转发路径的信息发送到所述SC；
    所述SC，还用于根据所述第一转发路径的信息和所述第二转发路径的信息将所述第一转发路径段和所述第二转发路径段连接形成所述转发路径。
PCT/CN2018/091688 2017-06-20 2018-06-15 一种网络中建立转发路径的方法、控制器及系统 Ceased WO2018233580A1 (zh)

Priority Applications (2)

Application Number Priority Date Filing Date Title
EP18821178.3A EP3618370B1 (en) 2017-06-20 2018-06-15 Method, controller and system for establishing forwarding path in network
US16/703,675 US11290568B2 (en) 2017-06-20 2019-12-04 Method, controller, and system for establishing forwarding path in network

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201710470662.9A CN107370673B (zh) 2017-06-20 2017-06-20 一种网络中建立转发路径的方法、控制器及系统
CN201710470662.9 2017-06-20

Related Child Applications (1)

Application Number Title Priority Date Filing Date
US16/703,675 Continuation US11290568B2 (en) 2017-06-20 2019-12-04 Method, controller, and system for establishing forwarding path in network

Publications (1)

Publication Number Publication Date
WO2018233580A1 true WO2018233580A1 (zh) 2018-12-27

Family

ID=60305528

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2018/091688 Ceased WO2018233580A1 (zh) 2017-06-20 2018-06-15 一种网络中建立转发路径的方法、控制器及系统

Country Status (4)

Country Link
US (1) US11290568B2 (zh)
EP (1) EP3618370B1 (zh)
CN (2) CN111654438B (zh)
WO (1) WO2018233580A1 (zh)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2021052248A1 (zh) * 2019-09-17 2021-03-25 中国移动通信有限公司研究院 链路资源发现方法、装置、设备及计算机可读存储介质

Families Citing this family (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111654438B (zh) 2017-06-20 2023-09-26 华为技术有限公司 一种网络中建立转发路径的方法、控制器及系统
CN114430390B (zh) 2017-12-29 2023-04-04 华为技术有限公司 一种用于获取跨域链路的方法及装置
CN109547347B (zh) * 2018-10-15 2020-12-25 浙江工业大学 一种多径通信网络化控制系统的路径调度与控制器协同设计方法
CN113259260B (zh) * 2020-02-11 2025-05-23 华为云计算技术有限公司 部署应用实例和调度应用实例的方法和装置
CN112511424B (zh) * 2020-05-15 2024-11-19 中兴通讯股份有限公司 转发和转发报文的方法、头节点、转发节点、存储介质
CN112566183B (zh) * 2020-11-20 2023-04-21 北京直真科技股份有限公司 一种自动开通5g传输电路的sdn控制器
CN114143004B (zh) * 2021-12-03 2023-07-28 网络通信与安全紫金山实验室 随机转发网络的部署方法、装置、设备及存储介质
CN114363404B (zh) * 2021-12-29 2024-11-22 凯通科技股份有限公司 一种混合组网下的以太网业务端到端还原方法及相关设备
JP7744606B2 (ja) * 2022-02-17 2025-09-26 Ｎｔｔ株式会社 通信システム、経路制御装置及び経路制御方法
US11792096B1 (en) * 2022-09-23 2023-10-17 Verizon Patent And Licensing Inc. Method and system for predictive and feedback management of network performance
US12495351B2 (en) * 2023-06-07 2025-12-09 Cisco Technology, Inc. Facilitating radio access network integration with data centers

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105960783A (zh) * 2013-12-26 2016-09-21 华为技术有限公司 域间sdn流量工程
CN106375214A (zh) * 2016-11-10 2017-02-01 北京邮电大学 一种基于sdn的层次化路由路径确定方法及装置
WO2017063694A1 (en) * 2015-10-14 2017-04-20 Telefonaktiebolaget Lm Ericsson (Publ) Control of multi-layer communication network
CN106789637A (zh) * 2015-11-24 2017-05-31 中国移动通信集团公司 一种跨域业务互通的路径建立方法、控制器及系统
CN107370673A (zh) * 2017-06-20 2017-11-21 华为技术有限公司 一种网络中建立转发路径的方法、控制器及系统

Family Cites Families (19)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US7496105B2 (en) * 2004-11-05 2009-02-24 Cisco Technology, Inc. System and method for retrieving computed paths from a path computation element using encrypted objects
US7649844B2 (en) * 2004-12-29 2010-01-19 Cisco Technology, Inc. In-order fibre channel packet delivery
US7492765B2 (en) * 2005-06-15 2009-02-17 Cisco Technology Inc. Methods and devices for networking blade servers
US7643430B2 (en) * 2005-11-30 2010-01-05 Cisco Technology, Inc. Methods and apparatus for determining reverse path delay
US20100125619A1 (en) * 2008-11-20 2010-05-20 Microsoft Corporation Deterministic selection of domain controllers in a multi-master database distributed directory service
CN103166876B (zh) * 2011-12-08 2018-05-04 中兴通讯股份有限公司 OpenFlow网络域间的数据发送方法及装置
US9360885B2 (en) * 2013-04-09 2016-06-07 International Business Machines Corporation Fabric multipathing based on dynamic latency-based calculations
CN104753713B (zh) * 2013-12-31 2019-02-05 华为技术有限公司 一种sdn部署业务的方法和sdn控制器
CN104104973B (zh) * 2014-06-12 2017-07-07 中山大学 一种应用于云媒体系统的群组带宽管理优化方法
CN104301391B (zh) * 2014-09-19 2019-02-22 北京邮电大学 多域光网络数据中心资源虚拟化映射方法
EP3229413B1 (en) * 2014-12-31 2021-11-17 Huawei Technologies Co., Ltd. Cross-domain cooperative method, cooperative device and control device for network as a service business
CN106302153B (zh) * 2015-05-11 2020-02-07 中兴通讯股份有限公司 多域控制器、单域控制器、软件定义光网络系统及方法
WO2017008285A1 (zh) * 2015-07-15 2017-01-19 华为技术有限公司 无线回程通信的方法和装置
US9667518B2 (en) * 2015-09-11 2017-05-30 Telefonaktiebolaget L M Ericsson (Publ) Method and system for delay measurement of a traffic flow in a software-defined networking (SDN) system
CN105515895B (zh) * 2015-11-30 2018-07-20 东南大学 基于软件定义网络的端到端路径上逐跳时延测量方法
CN106230512A (zh) * 2016-07-21 2016-12-14 国网信息通信产业集团有限公司 一种电力泛在光接入组网系统
CN106452958B (zh) * 2016-09-30 2020-04-07 中国联合网络通信集团有限公司 一种流量控制方法、系统及集中控制器
CN106506108A (zh) * 2016-12-06 2017-03-15 全球能源互联网研究院 基于软件定义网络sdn的同步报文调度方法及装置
CN106850435A (zh) * 2016-12-16 2017-06-13 北京格林伟迪科技股份有限公司 一种跨域端到端标签交换路径建立方法和系统

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105960783A (zh) * 2013-12-26 2016-09-21 华为技术有限公司 域间sdn流量工程
WO2017063694A1 (en) * 2015-10-14 2017-04-20 Telefonaktiebolaget Lm Ericsson (Publ) Control of multi-layer communication network
CN106789637A (zh) * 2015-11-24 2017-05-31 中国移动通信集团公司 一种跨域业务互通的路径建立方法、控制器及系统
CN106375214A (zh) * 2016-11-10 2017-02-01 北京邮电大学 一种基于sdn的层次化路由路径确定方法及装置
CN107370673A (zh) * 2017-06-20 2017-11-21 华为技术有限公司 一种网络中建立转发路径的方法、控制器及系统

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2021052248A1 (zh) * 2019-09-17 2021-03-25 中国移动通信有限公司研究院 链路资源发现方法、装置、设备及计算机可读存储介质

Also Published As

Publication number Publication date
CN107370673B (zh) 2020-04-28
US11290568B2 (en) 2022-03-29
CN107370673A (zh) 2017-11-21
EP3618370A4 (en) 2020-05-13
CN111654438A (zh) 2020-09-11
EP3618370A1 (en) 2020-03-04
US20200106858A1 (en) 2020-04-02
CN111654438B (zh) 2023-09-26
EP3618370B1 (en) 2021-09-22

Similar Documents

Publication Publication Date Title
WO2018233580A1 (zh) 一种网络中建立转发路径的方法、控制器及系统
US9838268B1 (en) Distributed, adaptive controller for multi-domain networks
US11528190B2 (en) Configuration data migration for distributed micro service-based network applications
EP3541028B1 (en) Adaptive load-balancing over a multi-point logical interface
US9819540B1 (en) Software defined network controller
US10200258B2 (en) Transaction integrity for network services configuration
CN103516602B (zh) 用于传输数据的方法、网络设备以及网络系统
US20180302286A1 (en) Network multi-source inbound quality of service methods and systems
JP6648892B2 (ja) 適合ネットワーク機能チェーン
US8897295B2 (en) Method and system for providing traffic engineering interworking
Bryskin et al. Policy-enabled path computation framework
WO2024036149A1 (en) Automated scaling of network topologies using unique identifiers
CN104994019B (zh) 一种用于sdn控制器的水平方向接口系统
CN107508730B (zh) 一种基于sdn网络的数据中心互联方法及装置
CN112436990B (zh) 一种数据转发方法、设备以及计算机可读存储介质
Casellas et al. Path Computation Elements (PCE): Applications and Status
CN217721212U (zh) 一种新型5g用sd-wan架构平台
Rahaman Rsvp and ldp: Protocols for traffic engineering
Huang et al. Nonuniform bandwidth reservation for tunnels in MPLS network using meter tables of OpenFlow
Binczewski et al. Enabling Service Market in Metro and Access Networks-The ADDONAS Project
Beyenel et al. Improving Quality of Service of Border Gateway Protocol Multiprotocol Label
WO2016023149A1 (zh) 一种业务服务质量保证方法与集中控制器
CN117255048A (zh) 确定业务传输策略的方法、装置及系统
Yabusaki et al. Network management for higher-bandwidth-use efficiency by optimizing paths in virtual networks
Asrat Improving Quality of Service of Border Gateway Protocol Multi protocol Label Switching Virtual Private Network of EthioTelecom Service Level Agreements

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: 18821178

Country of ref document: EP

Kind code of ref document: A1

ENP Entry into the national phase

Ref document number: 2018821178

Country of ref document: EP

Effective date: 20191128

NENP Non-entry into the national phase

Ref country code: DE