WO2016101469A1 - 一种基于捆绑接口的隧道带宽预留的方法及装置 - Google Patents

一种基于捆绑接口的隧道带宽预留的方法及装置 Download PDF

Info

Publication number
WO2016101469A1
WO2016101469A1 PCT/CN2015/077536 CN2015077536W WO2016101469A1 WO 2016101469 A1 WO2016101469 A1 WO 2016101469A1 CN 2015077536 W CN2015077536 W CN 2015077536W WO 2016101469 A1 WO2016101469 A1 WO 2016101469A1
Authority
WO
WIPO (PCT)
Prior art keywords
link
bandwidth
tunnel
interface
member link
Prior art date
Application number
PCT/CN2015/077536
Other languages
English (en)
French (fr)
Inventor
席媛媛
刘小飞
Original Assignee
中兴通讯股份有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 中兴通讯股份有限公司 filed Critical 中兴通讯股份有限公司
Publication of WO2016101469A1 publication Critical patent/WO2016101469A1/zh

Links

Images

Definitions

  • the present invention relates to the field of communications, and in particular to a method and apparatus for tunnel bandwidth reservation based on a bundle interface.
  • the Resource Reservation Protocol-Traffic Engineering (Resource Reservation Protocol-Traffic Engineering) is used in the current static routing/IP Radio Access Network (IP/ARP) environment.
  • the RSVP-TE tunnel is based on the bundled interface, such as the link aggregation (smart-group interface, SG interface for short) and pos-group interface (for example, PG interface).
  • the link state is unstable or the bundled interface or the link is faulty, the link is bound to each link in the bundled interface. If the link state is unstable, the bandwidth of the SG is shared. In the environment where the TE tunnel is deployed, but the link bandwidth is limited, all the TE tunnels share the bandwidth reserved by the SG.
  • the TE tunnel is frequently oscillated due to insufficient bandwidth resources, which has a large impact on service traffic.
  • a service tunnel such as a label forwarding protocol LSP tunnel, performs traffic forwarding and bandwidth reservation management based on the bundle interface as a whole, and the link state is unstable, the link bandwidth is limited, or the bundle interface and the link appear.
  • the bundled interface converges slowly.
  • each link on the bundle interface flaps which has a great impact on the forwarding of service traffic.
  • the service tunnels in the related technologies are configured to perform traffic forwarding and bandwidth reservation management based on the bundle interface.
  • the link state is unstable, the link bandwidth is limited, or the bundle interface and the link are faulty, the bundle interface itself converges slowly. It also causes the oscillation of each link in the bundled interface, which affects the forwarding of service traffic.
  • no effective solution has been proposed.
  • the embodiment of the invention provides a method and a device for tunnel bandwidth reservation based on a bundle interface, so as to at least solve the above problems in the related art.
  • a method for tunnel bandwidth reservation based on a bundle interface including: selecting a first member link of the bundle interface according to a preset selection policy of a member link of the bundle interface As a forwarding entry of the service tunnel, the protection relationship between the first member link and other valid member links of the bundle interface except the first member link.
  • the method further includes: performing bandwidth reservation on the service tunnel on the first member link.
  • the method further includes: switching traffic of the service tunnel from the first member link to the other active member link, where The other active member links share the traffic of the service tunnel in a manner of common load sharing.
  • the method further includes: selecting, according to the preset selection policy, a second member link as a forwarding entry of the forwarding entry of the service tunnel; forming a protection relationship between the second member link and other valid member links of the bundle interface except the second member link; Bandwidth reservation is performed on the service tunnel on the second member link.
  • the preset selection policy includes one of the following: if the unreserved bandwidth of the multiple member links meets the bandwidth requirement of the service tunnel, the bandwidth is not reserved in the multiple member links.
  • the minimum member link is the selected link; if the unreserved bandwidth of the multiple member links meets the bandwidth requirement of the service tunnel, the member link with the smallest index among the multiple member links is used as the member link a selected link; if the unreserved bandwidth of all the member links in the bundle interface does not satisfy the bandwidth requirement of the service tunnel, the one or more other services whose priority is lower than the service tunnel are removed. tunnel.
  • an apparatus for tunnel bandwidth reservation based on a bundle interface including: a first selection module, configured to select the according to a preset selection policy of a member link of a bundle interface
  • the first member link of the bundled interface is used as the outbound interface of the forwarding entry of the service tunnel;
  • the first forming module is configured to form the first member link and the bundled interface except the first member link. The protection relationship between other active member links.
  • the device further includes: a first reservation module, configured to perform bandwidth reservation on the service tunnel on the first member link.
  • the device further includes: a switching module, configured to switch traffic of the service tunnel from the first member link to the other valid member if the first member link fails A link, wherein the other effective member links share the traffic of the service tunnel by using a common load balancing manner.
  • a switching module configured to switch traffic of the service tunnel from the first member link to the other valid member if the first member link fails A link, wherein the other effective member links share the traffic of the service tunnel by using a common load balancing manner.
  • the device further includes: a second selection module, configured to select a second member link as a forwarding entry outbound interface of the service tunnel according to the preset selection policy; and a second forming module is configured to form a protection relationship between the second member link and other valid member links of the bundle interface except the second member link; and a second reservation module, configured to be in the second member chain The bandwidth of the service tunnel is reserved on the road.
  • a second selection module configured to select a second member link as a forwarding entry outbound interface of the service tunnel according to the preset selection policy
  • a second forming module is configured to form a protection relationship between the second member link and other valid member links of the bundle interface except the second member link
  • a second reservation module configured to be in the second member chain The bandwidth of the service tunnel is reserved on the road.
  • the preset selection policy includes at least one of the following: if the unreserved bandwidth of the multiple member links meets the bandwidth requirement of the service tunnel, the multiple member links are not reserved. The member link with the smallest bandwidth is the selected link; if the unreserved bandwidth of the multiple member links meets the bandwidth requirement of the service tunnel, the member link with the smallest index among the multiple member links is selected. As the selected link, if the unreserved bandwidth of all the member links in the bundle interface does not satisfy the bandwidth requirement of the service tunnel, the one or more others whose priority is lower than the service tunnel is removed. Business tunnel.
  • the first member link of the bundle interface is selected as the outbound interface of the forwarding entry of the service tunnel according to the preset selection policy of the member link of the bundle interface; the first member link and the first member link are formed.
  • the protection relationship between the active member links of the bundled interface except the first member link which solves the problem that the traffic tunnel existing in the related technology performs traffic forwarding and bandwidth reservation management based on the bundle interface as a whole.
  • the bandwidth is limited, or the bundled interface and the link are faulty, the bundled interface converges slowly.
  • the effect of the fast convergence of the bundled interface is achieved.
  • FIG. 1 is a flowchart of a tunnel bandwidth reservation method based on a bundle interface according to an embodiment of the present invention
  • FIG. 2 is a flowchart of another method for tunnel bandwidth reservation based on a bundle interface according to an embodiment of the present invention
  • FIG. 3 is a structural block diagram of a tunnel bandwidth reservation apparatus based on a bundle interface according to an embodiment of the present invention
  • FIG. 4 is a structural diagram of another tunnel bandwidth reservation apparatus based on a bundle interface according to an embodiment of the present invention.
  • FIG. 5 is a schematic diagram of networking of an RSVP-TE tunnel based on a bundle interface according to a preferred embodiment of the present invention
  • FIG. 6 is a schematic diagram of a bandwidth reservation of an RSVP-TE tunnel based on a bundle interface according to a preferred embodiment of the present invention
  • FIG. 7 is a flowchart of a SG-based TE-LSP tunnel bandwidth reservation method according to a preferred embodiment of the present invention.
  • FIG. 1 is a flowchart of a tunnel bandwidth reservation method based on a bundle interface according to an embodiment of the present invention. The process can include the following steps:
  • step S100 the first member link of the bundle interface is selected as the outgoing interface of the forwarding entry of the service tunnel according to the preset selection policy of the member link of the bundle interface;
  • the service tunnel in the device may be, for example, an RSVP-TE tunnel.
  • the current tunnel exit is a bundle interface such as a SG (smart-group interface) or a PG (pos-group interface)
  • traffic forwarding and bandwidth reservation management are performed based on the bundle interface as a whole.
  • the method of the present embodiment is to bind the service tunnel to the one or more member links of the bundle interface according to the pre-configured tunnel and the member link selection and selection policy in the bundle interface, as the forwarding entry of the tunnel. interface. From a macro perspective, the tunnel still forwards traffic based on the bundled interface and performs bandwidth reservation based on the bundled interface. Therefore, the service tunnel, such as the LSP itself, does not change its forwarding label.
  • Step S102 forming a protection relationship between the first member link and other valid member links of the bundle interface except the first member link.
  • the member link of the selected bundle interface becomes the first member link, which may be one link or several links.
  • the first member link and the tunnel form a binding relationship. From the perspective of traffic forwarding, the traffic of the tunnel is visible to the member link because the traffic of the tunnel is forwarded in the member link.
  • the other member links in the bundled interface form a protection relationship with the first member link. When the first member link fails, the service traffic on the link can be switched to other member links. The other member link temporarily forwards the process forwarding task it assumed before.
  • the first member link of the bundle interface is selected as the outbound interface of the forwarding entry of the service tunnel, and the first member link is formed by using the preset selection policy of the member link of the bundle interface.
  • the protection relationship between the active member links of the bundled interface except the first member link which solves the problem that the traffic tunnel existing in the related technology performs traffic forwarding and bandwidth reservation management based on the bundle interface as a whole. Belt When the bandwidth is limited, or the bundled interface and the link are faulty, the bundled interface converges slowly. This causes the link on the bundled interface to flap and affects the forwarding of service traffic.
  • a member link is selected to perform bandwidth reservation and traffic forwarding. When a bundled interface or some link is abnormal, you can quickly switch to other member links in the bundled interface to share the link. The road has no effect, effectively reducing the impact on traffic forwarding, and achieving fast convergence of the bundled interface.
  • FIG. 2 is a flowchart of another method for tunnel bandwidth reservation based on a bundle interface according to an embodiment of the present invention, as shown in FIG. 2 .
  • the process can include the following steps:
  • Step S200 and step S202 may refer to step S100 and step S102 shown in FIG. 1 respectively, and details are not described herein.
  • the method may further include:
  • Step S204 performing bandwidth reservation on the service tunnel on the first member link.
  • Bandwidth reservation refers to the application process of allocating bandwidth to users and network services. This includes assigning priorities to different traffic flows based on traffic flow importance and latency sensitivity. This maximizes the available bandwidth, and if the network is congested, low-priority services are discarded. Sometimes referred to as bandwidth allocation.
  • the service tunnel may microscopically reserve bandwidth based on the first member link.
  • the service tunnel is used to perform bandwidth reservation based on the member links in the bundle interface. Therefore, when the link is normal, the tunnel selects a member link to perform bandwidth reservation and traffic forwarding. If the link is abnormal, you can quickly switch to other member links in the bundled interface to perform the load balancing. This function has no effect on the member links that are in the normal state. This effectively reduces the impact on traffic forwarding and implements fast convergence of the bundled interface. Beneficial effect.
  • the method may further include: switching traffic of the service tunnel from the first member link to the other valid member A link, wherein the other effective member links share the traffic of the service tunnel by using a common load balancing manner.
  • the bundled interface is forwarded to the load balancing mode in the normal mode to perform temporary traffic protection.
  • a valid member link is a member link that does not have a fault in the bundle interface. It is feasible and feasible to perform traffic forwarding and bandwidth reservation based on a valid member link.
  • the common load sharing mode is related to the technology. The service tunnel is based on the traffic forwarding of the bundled interface, and bandwidth reservation is performed based on the bundle interface granularity.
  • the RSVP-TE tunnel and the SG bundle interface are used as an example.
  • the switch is quickly switched to other member interfaces of the SG.
  • different member interfaces are selected and different member interfaces are selected, when a member interface fails, different service tunnels do not affect each other.
  • the bandwidth reservation and traffic forwarding of the RSVP-TE tunnel are improved and optimized, and there is no impact on the protocol exchange and receiving packets of the original protocol layer.
  • the LSP of the TE tunnel can allocate resources to the LSP according to the bandwidth constraints of the SP and the load of the bundled member link.
  • the member link can be re-selected for the LSP when the member link fails. In this mechanism, the faulty link The LSP is not recalculated, and the fast convergence of the LSP is achieved, which has less impact on service traffic.
  • the method further includes: selecting, according to the preset selection policy, The two-member link serves as the outbound interface of the forwarding entry of the service tunnel; and the protection between the second member link and other effective member links of the bundle interface except the second member link Relationship; bandwidth reservation is performed on the service tunnel on the second member link.
  • Traffic forwarding and bandwidth reservation based on other member links can only be temporary. It is a temporary policy that the bundled interface converges quickly and cannot interfere with other active member links when the first member link fails.
  • the service tunnel of the first member link that has been bound is re-selected and assigned other valid member links, so that the member link forwarding bearer service traffic in the bundle interface is relatively stable and fixed. Operating mode.
  • the method may further include: detecting that the first member link failure disappears; and if the other valid member links cannot meet the bandwidth reservation of the service tunnel, according to the preset bundle interface. a member link selection policy, the third member link is selected as the outbound interface of the service tunnel forwarding entry; and the third member link is formed into a link protection relationship with other remaining member links in the bundle interface; The service tunnel reserves bandwidth based on the third member link.
  • the preset selection policy of the effective member in the tunnel selection binding interface may include at least:
  • the bandwidth of the active member interface does not meet the bandwidth requirements of the TE-LSP, you can determine whether the low-priority priority tunnel can be removed according to the established TE-LSP establishment priority and the established tunnel priority.
  • Priority preemption mode In this process, the tunnel state will be oscillated and the traffic will be interrupted. Or the high-priority tunnel will be removed, and the low-priority tunnel will be maintained. For example, if the low-priority tunnel is removed, the bandwidth is still not met. The bandwidth of the second-lowest priority tunnel is removed, and the remaining bandwidth is insufficient to establish a second-lowest priority tunnel. However, if a low-priority tunnel is established, the low-priority tunnel is preferentially established.
  • a device for tunnel bandwidth reservation based on a bundle interface is also provided.
  • the device is configured to implement the foregoing embodiments and preferred embodiments, and details are not described herein.
  • the term "module” may implement a combination of software and/or hardware of a predetermined function.
  • the apparatus described in the following embodiments is preferably implemented in software, hardware, or a combination of software and hardware, is also possible and contemplated.
  • FIG. 3 is a structural block diagram of a tunnel bandwidth reservation apparatus based on a bundle interface according to an embodiment of the present invention. As shown in FIG. 3, the apparatus may include a first selection module 30 and a first forming module 32. The device will be described below.
  • the first selection module 30 is configured to select a first member link of the bundle interface as a forwarding entry of the service tunnel according to a preset selection policy of the member link of the bundle interface; the first forming module 32 is connected to The first selection module 30 is configured to form a protection relationship between the first member link and other valid member links of the bundle interface except the first member link.
  • FIG. 4 is a structural block diagram of another tunnel bandwidth reservation apparatus based on a bundle interface according to an embodiment of the present invention. As shown in FIG. 4, the apparatus includes a first pre-fix, in addition to all the modules shown in FIG. The module 44 is left, and the device will be described below.
  • the first reservation module 44 is connected to the first forming module 42 and configured to perform bandwidth reservation on the service tunnel on the first member link.
  • the purpose of this embodiment is to use the preset selection policy of the member link of the bundle interface to select the first member link of the bundle interface as the outgoing interface entry of the service tunnel; forming the first member link and The protection relationship between the effective member links of the bundled interface except the first member link effectively solves the traffic forwarding and bandwidth reservation management of the service tunnel existing in the related technology based on the bundle interface as a whole.
  • the link bandwidth is limited, or the bundled interface and the link are faulty, the bundled interface converges slowly. This causes the link on the bundled interface to flap and affects the forwarding of service traffic.
  • the tunnel is fixed to select a member link for bandwidth reservation and traffic forwarding.
  • the sharing has no effect on the member links with normal status, effectively reducing the impact on traffic forwarding, and realizing the fast binding interface. Convergence benefits.
  • the apparatus further includes: a switching module, configured to: when the first member link fails, use the traffic of the service tunnel by the first The member link is switched to the other active member link, and the other effective member links share the traffic of the service tunnel in a manner of common load sharing.
  • a switching module configured to: when the first member link fails, use the traffic of the service tunnel by the first The member link is switched to the other active member link, and the other effective member links share the traffic of the service tunnel in a manner of common load sharing.
  • the apparatus further includes: a second selecting module, configured to select a second member link as a forwarding entry of the service tunnel according to the preset selection policy And a second forming module, configured to form a protection relationship between the second member link and other effective member links of the bundle interface except the second member link; and a second reserved module And setting a bandwidth reservation for the service tunnel on the second member link.
  • a second selecting module configured to select a second member link as a forwarding entry of the service tunnel according to the preset selection policy
  • a second forming module configured to form a protection relationship between the second member link and other effective member links of the bundle interface except the second member link
  • a second reserved module And setting a bandwidth reservation for the service tunnel on the second member link.
  • the apparatus further includes: a detecting module, configured to detect that the first member link fault disappears; and the third selecting module, configured to be in a link of other valid member links If the bandwidth reservation of the service tunnel is met, the third member link is selected as the outbound interface of the tunnel forwarding entry according to the preset selection policy of the member interface of the bundled interface; the third forming module is set to be The third member link forms a protection relationship between the links and other remaining member links in the bundle interface.
  • the third reservation module is configured to reserve bandwidth based on the third member link.
  • FIG. 5 is a schematic diagram of networking of an RSVP-TE tunnel based on a bundle interface according to a preferred embodiment of the present invention.
  • the SG has a new working mode, TE-TRUNK mode.
  • the main steps of the method for bandwidth reservation of the RSVP-TE tunnel based on the bundle interface described in the first example are as follows:
  • Step A Configure a dynamic RSVP-TE tunnel.
  • the outbound interface is the SG bundle interface. Set the bandwidth of the TE tunnel and set the available bandwidth of the tunnel of the outbound interface SG.
  • Step B A new working mode is added to the bundle interface, which is called TE-TRUNK mode.
  • the original load sharing mode is the normal mode.
  • the SG interface is enabled in TE-TRUNK mode (the default is normal mode).
  • Step C After the SG1 is enabled in the TE-TRUNK mode, the TE-LSP1 is bound to the link 1 according to the member interface selection policy, and the TE-LSP 2 and the link 2 form a binding relationship.
  • TE-LSP1 uses member port link 1 as the forwarding outbound interface and reserves bandwidth based on link 1.
  • TE-LSP2 uses member port link 2 as the forwarding outbound interface and reserves the band based on link 2.
  • the tunnel uses link 1 and link 2 for bandwidth reservation and bearer traffic forwarding.
  • a member link is selected as the outbound interface of the tunnel forwarding entry, and bandwidth reservation is performed based on the member link.
  • the member link and the SG interface form a protection relationship between the interfaces.
  • Step D When the link of the active member of the TE tunnel is faulty, the device quickly converges to the SG port to perform load balancing mode forwarding in the normal mode to perform temporary traffic protection.
  • the bandwidth reservation is based on the member port for bandwidth reservation, instead of bandwidth reservation based on the SG bundle interface granularity.
  • the link 1 and the other member links of the SG1, that is, the link 2 and the link 3 form a protection relationship;
  • Link 2 is in a protection relationship with other active member links of SG1, that is, link 1 and link 3.
  • Step E Re-converge to a new member link as the outbound interface of the tunnel forwarding entry based on the bandwidth of the configured TE tunnel and the bandwidth of other active member links in the SG interface, and perform bandwidth pre-emption on the new member link. stay.
  • the label forwarding of the LSP itself has no change.
  • FIG. 6 is a schematic diagram of RSVP-TE tunnel bandwidth reservation based on a bundle interface according to a preferred embodiment of the present invention.
  • TE-LSP1 when link 1 fails, TE-LSP1 quickly switches to other links in SG1 to forward traffic. Based on the member link selection policy, a new member link, link 3, is selected. Forward the outbound interface of the traffic and perform bandwidth reservation and traffic forwarding based on link 3. In this process, link 2 and TE-LSP 2 are not affected by link 1 failure, and TE LSP-1 only performs member interface update, and LSP does not need to be recalculated. From the perspective of SG1, SG1 achieves fast convergence and effectively reduces the impact on traffic forwarding of the SG interface.
  • FIG. 7 is a flow chart of a method for tunnel bandwidth reservation based on a bundle interface according to a preferred embodiment of the present invention.
  • Step S700 Configure a bundle interface SG and configure an RSVP-TE tunnel.
  • the member interfaces are link-1 (link-1), link 2 (link-2), and link 3 (link-3).
  • the tunnel forwarding outbound interface is SG1, and the tunnel bandwidth is 10000 and 15000 respectively.
  • the numbers are used as examples only, which does not represent the limitation of this example.
  • step S702 the SG is configured to be in the TE-TRUNK mode. Enable TE-TRUNK mode on the SG1 interface.
  • Step S704 According to the member port selection policy, select a member interface of the SG as the outgoing interface of the TE tunnel, reserve bandwidth on the member interface, and form protection between the member interface and the SG interface.
  • TE1 uses member port link 1 as the forwarding outbound interface and reserves bandwidth based on link 1.
  • TE2 uses member port link 2 as the forwarding outbound interface and reserves bandwidth based on link 2 (when SG When working in TE-TRUNK mode, bandwidth reservation is based on member ports for bandwidth reservation, instead of bandwidth reservation based on SG bundle interface granularity.
  • Member interface selection can include the following strategies:
  • the bandwidth of the active member interface does not meet the bandwidth requirements of the TE LSP, you can determine whether the tunnel with the lower priority can be removed according to the establishment priority of the new TELSP and the retention priority of the created tunnel. Mode), in this process, the tunnel state will be oscillated, the traffic will be interrupted; or the high-priority tunnel will be removed, and the low-priority tunnel will be maintained. For example, after the low-priority tunnel is removed, the bandwidth is still not satisfied. The bandwidth of the low-priority tunnel is satisfied. The remaining bandwidth is insufficient to establish the second-lowest priority tunnel. However, if the low-priority tunnel is established, the low-priority tunnel is preferentially established.
  • step S706 when the member interface link is faulty, the traffic is quickly switched to the SG port to perform load sharing and forwarding in the normal mode.
  • the link 1 and the SG1 form a protection relationship for the TE1 egress link 1.
  • the link 2 and the SG1 form a protection relationship.
  • the link of the link 1 of the outbound interface of the TE1 tunnel is faulty.
  • the device quickly switches to the SG1 interface and temporarily uses the SG1 load to share the forwarding traffic.
  • Step S708 According to the member interface selection policy, the TE tunnel converges to the new member interface, reserves bandwidth on the new member interface, and forms an interface protection relationship with the SG interface.
  • TE1 converges to the member port link-3 as the forwarding outbound interface and reserves bandwidth based on link-3.
  • link-3 and SG1 form a protection relationship.
  • the TE tunnel recalculates the new member interface, in the new The bandwidth is reserved on the member interface and forms an interface protection relationship with the SG interface.
  • the method further includes: according to the member interface selection policy, the TE1 reuses the member port link 1 as the forwarding outbound interface, and reserves the bandwidth based on the link 1. At the forwarding level, link 1 and SG1 form a protection relationship again.
  • a member interface is selected to perform bandwidth reservation and traffic forwarding.
  • the link is quickly switched to other member links of the bundle interface.
  • the bandwidth reservation and traffic forwarding of the tunnel are improved and optimized, and there is no impact on the protocol exchange and receiving packets of the original protocol layer.
  • the tunnel can allocate resources to the tunnel according to the bandwidth constraints and the bundled member link load.
  • the member link can be re-selected for the service tunnel when the member link fails. In this mechanism, the LSP of the faulty link does not. Recalculate to achieve fast convergence of LSPs and have less impact on service traffic.
  • modules or steps of the present invention described above can be implemented by a general-purpose computing device that can be centralized on a single computing device or distributed across a network of multiple computing devices. Alternatively, they may be implemented by program code executable by the computing device such that they may be stored in the storage device by the computing device and, in some cases, may be different from the order herein. Perform the steps shown or described, or separate them into individual integrated circuit modules, or Multiple of these modules or steps are fabricated as a single integrated circuit module. Thus, the invention is not limited to any specific combination of hardware and software.
  • the method and apparatus for tunnel bandwidth reservation based on the bundle interface provided by the embodiments of the present invention have the following beneficial effects: the tunnel bandwidth can be ensured, and the interface bandwidth utilization is improved. Moreover, only the bandwidth reservation and traffic forwarding of the tunnel are improved and optimized, and the protocol exchange and receiving packets of the original protocol layer have no effect.
  • the tunnel can allocate resources to the tunnel according to the bandwidth constraint and the bundled member link load, and can re-select the member link for the service tunnel when the member link fails. In this mechanism, the LSP of the faulty link It does not recalculate and achieves fast convergence of LSPs, which has less impact on service traffic.

Landscapes

  • Data Exchanges In Wide-Area Networks (AREA)

Abstract

本发明公开了一种基于捆绑接口的隧道带宽预留的方法及装置，其中，该方法包括：根据捆绑接口的成员链路的预设选择策略，选择所述捆绑接口的第一成员链路作为业务隧道的转发表项出接口；形成所述第一成员链路与所述捆绑接口中除所述第一成员链路之外的其他有效成员链路之间的保护关系，有效地解决了相关技术中存在的业务隧道基于捆绑接口整体进行流量转发以及带宽预留管理，在链路带宽有限或者捆绑接口和链路出现故障时，捆绑接口自身收敛较慢，并会造成捆绑接口中每一条链路出现震荡，影响业务流量的转发的问题，从而有效地减少了对流量转发的影响，实现了捆绑接口的快速收敛。

Description

一种基于捆绑接口的隧道带宽预留的方法及装置 技术领域
本发明涉及通信领域，具体而言，涉及一种基于捆绑接口的隧道带宽预留的方法及装置。
背景技术
当前静态路由/无线接入网IP化(Static Router/IP Radio Access Network，简称为SR/IPRAN)的环境下，设备中基于流量工程扩展的资源预留协议(Resource Reservation Protocol-Traffic Engineering，简称为RSVP-TE)隧道出口为链路聚合(Link Aggregation)的智慧群组(smart-group接口，简称为SG接口)、pos-group接口(简称为PG接口)等捆绑接口时，均是基于捆绑接口整体进行流量转发以及带宽预留管理。在没有部署流量工程(Traffic Engineering，简称为TE)隧道保护的环境中，如果链路状态不稳定或者捆绑接口、链路出现故障时时，依靠捆绑接口自身的收敛，捆绑接口中每一条链路都出收敛出现震荡，会对业务流量造成较大的影响；在部署较多TE隧道，但是链路带宽有限的环境中，所有TE隧道共享SG整体预留的带宽，如果链路状态不稳定时，也会频繁造成TE隧道由于带宽资源不足而震荡，对业务流量造成较大的影响。
综上所述，在相关技术中，业务隧道例如标签转发协议LSP隧道基于捆绑接口整体进行流量转发以及带宽预留管理，在链路状态不稳定、链路带宽有限、或者捆绑接口和链路出现故障时，捆绑接口自身收敛较慢，并会造成捆绑接口中每一条链路出现震荡，对业务流量的转发带来很大的影响。
针对相关技术中存在的业务隧道基于捆绑接口整体进行流量转发以及带宽预留管理，在链路状态不稳定、链路带宽有限、或者捆绑接口和链路出现故障时，捆绑接口自身收敛较慢，并会造成捆绑接口中每一条链路出现震荡，影响业务流量的转发的问题，目前尚未提出有效的解决方案。
发明内容
本发明实施例提供了一种基于捆绑接口的隧道带宽预留的方法及装置，以至少解决相关技术中存在的上述问题。
根据本发明实施例的一个方面，提供了一种基于捆绑接口的隧道带宽预留的方法，包括：根据捆绑接口的成员链路的预设选择策略，选择所述捆绑接口的第一成员链路作为业务隧道的转发表项出接口；形成所述第一成员链路与所述捆绑接口中除所述第一成员链路之外的其他有效成员链路之间的保护关系。
进一步地，所述方法还包括：在所述第一成员链路上对所述业务隧道进行带宽预留。
进一步地，在所述第一成员链路出现故障的情况下，所述方法还包括：将所述业务隧道的流量由所述第一成员链路切换到所述其他有效成员链路，其中，所述其他有效成员链路通过普通负荷分担的方式分担所述业务隧道的流量。
进一步地，将所述业务隧道的流量由所述第一成员链路切换到所述其他有效成员链路之后，所述方法还包括：根据所述预设选择策略，选择第二成员链路作为所述业务隧道的转发表项出接口；形成所述第二成员链路与所述捆绑接口中除所述第二成员链路之外的其他有效成员链路之间的保护关系；在所述第二成员链路上对所述业务隧道进行带宽预留。
进一步地，所述预设选择策略包括以下之一：在多条成员链路的未预留带宽满足所述业务隧道的带宽需求的情况下，将所述多条成员链路中未预留带宽最小的成员链路作为被选择链路；在多条成员链路的未预留带宽满足所述业务隧道的带宽需求的情况下，将所述多条成员链路中索引最小的成员链路作为被选择链路；在所述捆绑接口中所有成员链路的未预留带宽均不满足所述业务隧道的带宽需求的情况下，拆除优先级低于所述业务隧道的一个或多个其他业务隧道。
根据本发明实施例的另一方面，提供了一种基于捆绑接口的隧道带宽预留的装置，包括：第一选择模块，设置为根据捆绑接口的成员链路的预设选择策略，选择所述捆绑接口的第一成员链路作为业务隧道的转发表项出接口；第一形成模块，设置为形成所述第一成员链路与所述捆绑接口中除所述第一成员链路之外的其他有效成员链路之间的保护关系。
进一步地，所述装置还包括：第一预留模块，设置为在所述第一成员链路上对所述业务隧道进行带宽预留。
进一步地，所述装置还包括：切换模块，设置为在所述第一成员链路出现故障的情况下，将所述业务隧道的流量由所述第一成员链路切换到所述其他有效成员链路，其中，所述其他有效成员链路通过普通负荷分担的方式分担所述业务隧道的流量。
进一步地，所述装置还包括：第二选择模块，设置为根据所述预设选择策略，选择第二成员链路作为所述业务隧道的转发表项出接口；第二形成模块，设置为形成所述第二成员链路与所述捆绑接口中除所述第二成员链路之外的其他有效成员链路之间的保护关系；第二预留模块，设置为在所述第二成员链路上对所述业务隧道进行带宽预留。
进一步地，所述预设选择策略包括以下至少之一：在多条成员链路的未预留带宽满足所述业务隧道的带宽需求的情况下，将所述多条成员链路中未预留带宽最小的成员链路作为被选择链路；在多条成员链路的未预留带宽满足所述业务隧道的带宽需求的情况下，将所述多条成员链路中索引最小的成员链路作为被选择链路；在所述捆绑接口中所有成员链路的未预留带宽均不满足所述业务隧道的带宽需求的情况下，拆除优先级低于所述业务隧道的一个或多个其他业务隧道。
通过本发明，采用根据捆绑接口的成员链路的预设选择策略，选择所述捆绑接口的第一成员链路作为业务隧道的转发表项出接口；形成所述第一成员链路与所述捆绑接口中除所述第一成员链路之外的其他有效成员链路之间的保护关系，解决了相关技术中存在的业务隧道基于捆绑接口整体进行流量转发以及带宽预留管理，在链路带宽有限、或者捆绑接口和链路出现故障时，捆绑接口自身收敛较慢，并会造成捆绑接口中每一条链路出现震荡，影响业务流量的转发的问题，，从而有效地减少了对流量转发的影响，实现了捆绑接口的快速收敛。
附图说明
此处所说明的附图用来提供对本发明的进一步理解，构成本申请的一部分，本发明的示意性实施例及其说明用于解释本发明，并不构成对本发明的不当限定。在附图中：
图1是根据本发明实施例的一种基于捆绑接口的隧道带宽预留方法的流程图；
图2是根据本发明实施例的另一种基于捆绑接口的隧道带宽预留方法的流程图；
图3是根据本发明实施例的一种基于捆绑接口的隧道带宽预留装置的结构框图；
图4是根据本发明实施例的另一种基于捆绑接口的隧道带宽预留装置的结构图；
图5是本发明优选实施例基于捆绑接口的RSVP-TE隧道组网示意图；
图6是本发明优选实施例基于捆绑接口的RSVP-TE隧道带宽预留的示意图；
图7是根据本发明优选实施例的一种基于SG的TE-LSP隧道带宽预留方法的流程图。
具体实施方式
下文中将参考附图并结合实施例来详细说明本发明。需要说明的是，在不冲突的情况下，本申请中的实施例及实施例中的特征可以相互组合。
在本实施例中提供了一种基于捆绑接口的隧道带宽预留方法，图1是根据本发明实施例的一种基于捆绑接口的隧道带宽预留方法的流程图，如图1所示，该流程可以包括如下步骤：
步骤S100，根据捆绑接口的成员链路的预设选择策略，选择所述捆绑接口的第一成员链路作为业务隧道的转发表项出接口；
SR/IPRAN即无线接入网IP化的环境下，设备中的业务隧道，例如可是RSVP-TE隧道。相关技术中，当前隧道出口为SG(smart-group接口)、PG(pos-group接口)等捆绑接口时，均是基于捆绑接口整体进行流量转发以及带宽预留管理。本实施例提供方法根据预先设置的隧道与捆绑接口中成员链路选择选择策略，将业务隧道基于捆绑接口中的某一个或多个成员链路进行绑定，作为所述隧道的转发表项出接口。从宏观上来看，该隧道依然是基于该捆绑接口进行转发流量并基于捆绑接口进行带宽预留，因此业务隧道例如LSP其本身转发标签无变化。
步骤S102，形成所述第一成员链路与所述捆绑接口中除所述第一成员链路之外的其他有效成员链路之间的保护关系。
经过步骤S100，被选择的捆绑接口中成员链路，成为第一成员链路，可以是一条链路，也可以是若干链路。第一成员链路和隧道之间与形成捆绑关系，从流量转发的角度，隧道的业务流量对于该成员链路是可见，因为隧道的流量是在该成员链路中进行转发。并且捆绑接口中的其他成员链路与第一成员链路形成了捆绑接口内的保护关系，即第一成员链路出现故障时，该链路上的业务流量可以切换指其他成员链路，由其他成员链路暂时转发它之前承担的流程转发任务。
通过上述步骤，采用根据捆绑接口的成员链路的预设选择策略，选择所述捆绑接口的第一成员链路作为业务隧道的转发表项出接口；形成所述第一成员链路与所述捆绑接口中除所述第一成员链路之外的其他有效成员链路之间的保护关系，解决了相关技术中存在的业务隧道基于捆绑接口整体进行流量转发以及带宽预留管理，在链路带 宽有限、或者捆绑接口和链路出现故障时，捆绑接口自身收敛较慢，并会造成捆绑接口中每一条链路出现震荡，影响业务流量的转发的问题，，达到了在链路正常情况下隧道固定选择某个成员链路进行带宽预留与流量转发，当捆绑接口或某些链路异常时，可以快速切换到捆绑接口中其他状态正常的成员链路进行分担，对于状态正常的成员链路没有影响，有效地减少了对流量转发的影响，实现了捆绑接口快速收敛的。
在本实施例中还提供了一种基于捆绑接口的隧道带宽预留方法，图2是根据本发明实施例的另一种基于捆绑接口的隧道带宽预留方法的流程图，如图2所示，该流程可以包括如下步骤：
步骤S200、步骤S202可以分别参考图1所示的步骤S100、步骤S102，在此不赘述。在步骤S202之后，还可以包括：
步骤S204，在所述第一成员链路上对所述业务隧道进行带宽预留。
带宽预留是指把带宽分配给用户和网络服务的应用程序过程。包括根据业务流重要性和延迟敏感性为不同业务流分配优先级。这样就最大限度地利用了可用带宽，如果网络发生拥塞，低优先级的业务就被丢弃。有时也称为带宽分配。根据步骤S200中选择的第一成员链路，业务隧道从微观上可以基于第一成员链路进行带宽预留。
通过上述步骤，采用业务隧道基于捆绑接口中的成员链路进行带宽预留，从而实现了在链路正常情况下隧道固定选择某个成员链路进行带宽预留与流量转发，当捆绑接口或某些链路异常时，可以快速切换到捆绑接口中其他状态正常的成员链路进行分担，对于状态正常的成员链路没有影响，有效地减少了对流量转发的影响，实现了捆绑接口快速收敛的有益效果。
在优选的实施方式中，在所述第一成员链路出现故障的情况下，所述方法还可以包括：将所述业务隧道的流量由所述第一成员链路切换到所述其他有效成员链路，其中，所述其他有效成员链路通过普通负荷分担的方式分担所述业务隧道的流量。
在业务隧道流量的转发过程中，绑定的第一成员链路出现故障的情况下，快速收敛到捆绑接口进行普通模式的负荷分担模式转发，进行暂时的流量保护。有效成员链路是指捆绑接口中不存在故障的成员链路，基于有效的成员链路进行流量转发和带宽预留才是可行和可行的。普通负荷分担模式即是相关技术中，业务隧道基于捆绑接口的流量转发，并基于捆绑接口粒度进行带宽预留。
本实施例的目的，以RSVP-TE隧道和SG捆绑接口为例说，当绑定的成员链路异常时，通过快速切换到SG其他成员口进行分担。走相同SG口的不同隧道，选择不同成员接口的情况下，某成员接口故障时，不同业务隧道间相互不影响。仅对RSVP-TE隧道的带宽预留和流量转发进行改进优化，对原有协议层面的协议交互收发包无影响。TE隧道的LSP能够根据SP的带宽约束条件以及捆绑成员链路负载合理分配资源给LSP，同时可以在成员链路故障的时候能够为LSP重新选择成员链路，在这种机制下，故障链路的LSP不会重新计算，达到LSP的快速收敛，对业务流量的影响较小。
在优选的实施方式中，在将所述业务隧道的流量由所述第一成员链路切换到所述其他有效成员链路之后，所述方法还包括：根据所述预设选择策略，选择第二成员链路作为所述业务隧道的转发表项出接口；形成所述第二成员链路与所述捆绑接口中除所述第二成员链路之外的其他有效成员链路之间的保护关系；在所述第二成员链路上对所述业务隧道进行带宽预留。基于其他成员链路进行流量转发及带宽预留只能是暂时，只是对第一成员链路出现故障时，捆绑接口快速收敛，不能对其他有效成员链路造成干扰的临时性的策略。在预定的时间内，必须对失去绑定的第一成员链路的业务隧道重新选择和分配其他的有效成员链路，使捆绑接口中的成员链路转发承载业务流量处于一个相对稳定和固定的工作模式。
在优选的实施方式中，所述方法还可以包括：检测到所述第一成员链路故障消失；在其他有效成员链路无法满足业务隧道的带宽预留的情况下，根据预设的捆绑接口成员链路的选择策略，选择第三成员链路作为所述业务隧道转发表项出接口；将所述第三成员链路与捆绑接口里的其他剩余成员链路形成链路间的保护关系；所述业务隧道基于所述第三成员链路预留带宽。
在第一成员链路存在的故障的情况下，业务隧道的流量转发切换到所述其他有效成员链路。对其他有效成员链路也是一种负担，而且有可能会拆掉其他有效成员链路上的业务，因此基于其他有效成员链路的流量转发及带宽预留是暂时的。当原先存在故障的第一成员链路恢复正常工作时，出于成员链路负载均衡的原则，有必要进行一次新的隧道选择成员链路。当然在其他有效成员链路能够满足业务隧道的带宽预留、已经形成稳定的流量转发、带宽预留的情况下，没有必要再次选择第一成员链路。
在上述实施例中，隧道选择捆绑接口中的有效成员的预设选择策略至少可以包括：
1、当多条成员链路满足隧道的带宽需求的情况下，选择未预留带宽最小的成员链路，即在多条成员链路的未预留带宽满足所述业务隧道的带宽需求的情况下，将所述多条成员链路中未预留带宽最小的成员链路作为被选择链路；
2、当存在多条满足条件的成员链路的情况下，选择索引最小的成员链路，即在多条成员链路的未预留带宽满足所述业务隧道的带宽需求的情况下，将所述多条成员链路中索引最小的成员链路作为被选择链路；
3、当所述捆绑接口中所有成员链路带宽均不满足隧道的带宽需求时，对比新建隧道与已创建隧道的优先级，确定拆掉低优先级的隧道，即在所述捆绑接口中所有成员链路的未预留带宽均不满足所述业务隧道的带宽需求的情况下，拆除优先级低于所述业务隧道的一个或多个其他业务隧道。
在上述预设选择策略中，对于优先级相同的业务隧道，采用先申请，先建立的原则。
下面以标签转发协议LSP隧道和捆绑接口SG为例说明，选择捆绑接口中成员链路的策略如下：
1.当有多条活动成员接口满足TE-LSP的带宽需求时，选择未预留带宽最小的成员接口，以有效的利用带宽，减少带宽碎片；
2.当所有活动成员接口带宽均不满足TE-LSP的带宽需求时，需要根据新建TE-LSP的建立优先级与已创建隧道的保持优先级决定能否拆掉低保持优先级的隧道(支持优先级抢占方式)，在此过程中隧道状态会有震荡，流量中断；或者出现高优先级的隧道拆掉，低优先级的隧道保持，比如拆掉低优先级的隧道之后带宽仍不满足，拆掉次低优先级隧道带宽满足，此时剩余的带宽又不足以建立次低优先级隧道，但是可以满足建立低优先级隧道，则优先建立低优先级隧道；
3.对于优先级相同的隧道，则采用先申请，先建立的原则；
4.对以上原则，当存在多条满足条件的成员接口时，选择索引最小的成员接口。
基于负载均衡的原则，当TE1、TE2隧道的出口为同一个捆绑接口例如SG1时，不同的隧道会尽可能选择不同的成员端口进行带宽预留，结果可能使得多数情况下，多个成员链路会有一定的空闲容量，当某个成员链路发生故障，那么其它成员链路的空闲资源可以为其提供保护，也合理提高了带宽利用率。
在优选的实施方式中，为了尽量在有效链路生存期内始终保持均衡负载，从而在故障时更有效地保护LSP隧道，需要在下面情况发生时，对链路当前承载的LSP，判断是否需要调整其成员链路位置，从而进行LSP带宽调整：捆绑接口中成员链路故障消失；成员链路添加或减少；LSP删除或增加等。
在本实施例中还提供了一种基于捆绑接口的隧道带宽预留的装置，该装置设置为实现上述实施例及优选实施方式，已经进行过说明的不再赘述。如以下所使用的，术语“模块”可以实现预定功能的软件和/或硬件的组合。尽管以下实施例所描述的装置较佳地以软件来实现，但是硬件，或者软件和硬件的组合的实现也是可能并被构想的。
图3是根据图3是根据本发明实施例的一种基于捆绑接口的隧道带宽预留装置的结构框图，如图3所示，该装置可以包括第一选择模块30和第一形成模块32，下面对该装置进行说明。
第一选择模块30，设置为根据捆绑接口的成员链路的预设选择策略，选择所述捆绑接口的第一成员链路作为业务隧道的转发表项出接口；第一形成模块32，连接至第一选择模块30，设置为形成所述第一成员链路与所述捆绑接口中除所述第一成员链路之外的其他有效成员链路之间的保护关系。
图4是是根据本发明实施例的另一种基于捆绑接口的隧道带宽预留装置的结构框图，如图4所示，该装置除包括图3所示的所有模块外，还包括第一预留模块44，下面对该装置进行说明。
第一预留模块44，连接至第一形成模块42，设置为在所述第一成员链路上对所述业务隧道进行带宽预留。
本实施例的目的是采用根据捆绑接口的成员链路的预设选择策略，选择所述捆绑接口的第一成员链路作为业务隧道的转发表项出接口；形成所述第一成员链路与所述捆绑接口中除所述第一成员链路之外的其他有效成员链路之间的保护关系，有效地解决了相关技术中存在的业务隧道基于捆绑接口整体进行流量转发以及带宽预留管理，在链路带宽有限、或者捆绑接口和链路出现故障时，捆绑接口自身收敛较慢，并会造成捆绑接口中每一条链路出现震荡，对业务流量的转发带来很大影响的问题，达到了在链路正常情况下隧道固定选择某个成员链路进行带宽预留与流量转发，当捆绑接口或某些链路异常时，可以快速切换到捆绑接口中其他状态正常的成员链路进行分担，对于状态正常的成员链路没有影响，有效地减少了对流量转发的影响，实现了捆绑接口快速收敛的有益效果。
在本实施例中，可选地实施方式是，所述装置还包括：切换模块，设置为在所述第一成员链路出现故障的情况下，将所述业务隧道的流量由所述第一成员链路切换到所述其他有效成员链路，其中，所述其他有效成员链路通过普通负荷分担的方式分担所述业务隧道的流量。
在本实施例中，可选地实施方式是，所述装置还包括：第二选择模块，设置为根据所述预设选择策略，选择第二成员链路作为所述业务隧道的转发表项出接口；第二形成模块，设置为形成所述第二成员链路与所述捆绑接口中除所述第二成员链路之外的其他有效成员链路之间的保护关系；第二预留模块，设置为在所述第二成员链路上对所述业务隧道进行带宽预留。
在本实施例中，可选地实施方式是，所述装置还包括：检测模块，设置为检测到所述第一成员链路故障消失；第三选择模块，设置为在其他有效成员链路无法满足所述业务隧道的带宽预留的情况下，根据预设的捆绑接口成员链路的选择策略，选择第三成员链路作为所述隧道转发表项出接口；第三形成模块，设置为将所述第三成员链路与捆绑接口里的其他剩余成员链路形成链路间的保护关系；第三预留模块，设置为所述隧道基于所述第三成员链路预留带宽。
上述在此需要注意的是，第一、二、三选择模块，第一、二、三预留模块，第一、二、三形成模块，由于在它们的功能是相同或相似的，因此可以是同一个具有该功能的实体装置或虚拟模块，也可以是不同的实体装置或虚拟模块。
下面再结合具体实施例对本发明进行说明。
实例一：
图5是本发明优选实施例基于捆绑接口的RSVP-TE隧道组网示意图。
如图5中的上图所示，PE1、PE2两台设备的组网。两设备之间有3条物理链路，分别为链路1(link-1)、链路2(link-2)、链路3(link-3)，这3条物理链路形成一个捆绑接口SG1。即捆绑接口SG中有三条有效成员链路即没有故障，分别是链路1、链路2、链路3。基于捆绑接口进行转发流量和带宽预留的业务隧道有TE-LSP1、TE-LSP2。这样一来就形成了两台PE设备之间有2条RSVP-TE隧道TE LSP-1和TE LSP-2，出接口均为SG1。
根据本实例提供的方法，SG有一种新的工作模式即TE-TRUNK模式。本实例一所述的基于捆绑接口的RSVP-TE隧道带宽预留的方法实现主要步骤如下：
步骤A.配置动态RSVP-TE隧道，出接口为SG捆绑接口。配置TE隧道带宽，配置出接口SG的隧道可用带宽。
步骤B.捆绑接口新增一种工作模式，称之为TE-TRUNK模式，原来默认负荷分担模式普通模式。SG接口开启TE-TRUNK模式(默认为普通模式)。
步骤C.将SG1开启TE-TRUNK模式后，根据成员接口选择策略，TE-LSP1与链路1形成绑定关系，TE-LSP2与链路2形成绑定关系。TE-LSP1使用成员端口链路1作为转发出接口，并基于链路1预留带宽；TE-LSP2使用成员端口链路2作为转发出接口，并基于链路2预留带。隧道分别使用链路1和链路2进行带宽预留和承接流量转发。
根据配置的TE隧道的带宽和SG口中各活动成员接口的带宽，选取一个成员链路作为该隧道转发表项出接口，基于该成员链路进行带宽预留。将成员链路与SG口形成接口间的保护关系。
步骤D.TE隧道有效成员链路故障时，快速收敛到SG口进行普通模式的负荷分担模式转发，进行暂时的流量保护。
此时带宽预留为基于成员端口进行带宽预留，而不是基于SG捆绑接口粒度进行带宽预留。在流量转发层面，对于TE-LSP1转发出接口链路1，将链路1与SG1其他成员链路即链路2和链路3形成保护关系；对于TE-LSP2的转发出接口链路2，将链路2与SG1其他有效成员链路即链路1、链路3形成保护关系。
步骤E.根据配置的TE隧道的带宽和SG口中其他活动成员链路的带宽，重新收敛到一个新的成员链路作为该隧道转发表项出接口，并在新的成员链路上进行带宽预留。LSP本身标签转发无变化。
图6是本发明优选实施例基于捆绑接口的RSVP-TE隧道带宽预留的示意图。
如图6所示，当链路1出现故障时，TE-LSP1迅速切换至SG1中其他链路中进行转发流量，并基于成员链路选择策略，选择了新的成员链路即链路3作为转发流量的出接口链路并基于链路3进行带宽预留和流量转发。在此过程中，链路2和TE-LSP2并没有收到链路1故障的影响，而TE LSP-1也仅进行成员接口的更新，LSP不需要重新计算。从SG1来看，SG1实现了快速收敛，有效地减少了对SG接口的流量转发的影响。
实例二：
图7是根据本发明优选实施例的一种基于捆绑接口的隧道带宽预留方法的流程图。为了让本领域的技术人员更好的理解本发明的方法，下面结合附图7进一步描述本发明的具体实施方式。
步骤S700，配置捆绑接口SG，配置RSVP-TE隧道。
在设备上配置捆绑接口SG1，其中成员接口分别为链路1(link-1)、链路2(link-2)、链路3(link-3)。配置SG1接口可供动态隧道使能的带宽。配置两条RSVP TE隧道TE-LSP1，TE-LSP2，隧道转发出接口均为SG1，并分别配置隧道带宽10000，15000，数字仅做示例用，不代表对本实例的限制。
步骤S702，配置SG为TE-TRUNK模式。在SG1接口下开启TE-TRUNK模式。
步骤S704，根据成员端口选择策略，选取SG中某成员接口作为TE隧道的转发出接口，在该成员接口上预留带宽,将该成员接口与SG口形成接口间的保护。
根据成员接口选择策略，TE1使用成员端口链路1作为转发出接口，并基于链路1预留带宽；TE2使用成员端口链路2作为转发出接口，并基于链路2预留带宽(当SG工作在TE-TRUNK模式时，此时带宽预留为基于成员端口进行带宽预留，而不是基于SG捆绑接口粒度进行带宽预留)。
成员接口选择可以包括如下策略：
1.当有多条活动成员接口满足TE LSP的带宽需求时，选择未预留带宽最小的成员接口，以有效的利用带宽，减少带宽碎片；当存在多条满足条件的成员接口时，选择索引最小的成员接口；
2.当所有活动成员接口带宽均不满足TE LSP的带宽需求时，需要根据新建TELSP的建立优先级与已创建隧道的保持优先级决定能否拆掉低保持优先级的隧道(支持优先级抢占方式)，在此过程中隧道状态会有震荡，流量中断；或者出现高优先级的隧道拆掉，低优先级的隧道保持，比如拆掉低优先级的隧道之后带宽仍不满足，拆掉次低优先级隧道带宽满足，此时剩余的带宽又不足以建立次低优先级隧道，但是可以满足建立低优先级隧道，则优先建立低优先级隧道；
3.对于优先级相同的隧道，则采用先申请，先建立的原则。
基于负载均衡的原则，当TE1、TE2隧道的出口为同一个SG1时，不同的隧道会尽可能选取不同的成员端口进行带宽预留，结果可能使得多数情况下，多个成员链路会有一定的空闲容量，当某个成员链路发生故障，那么其它成员链路的空闲资源可以为其提供保护，也合理提高了带宽利用率。
步骤S706，在成员接口链路故障时，流量快速切换到SG口进行普通模式的负荷分担转发。
转发层面，对于TE1出口链路1，将链路1与SG1形成保护关系；对于TE2出口链路2，将链路2与SG1形成保护关系。TE1隧道转发出接口链路1链路故障，设备快速切换到SG1接口，暂时使用SG1负荷分担转发流量。
步骤S708，根据成员接口选择策略，TE隧道收敛到新的成员接口上，在新的成员接口上预留带宽，并将其与SG口形成接口保护关系。
根据成员接口选择策略，TE1收敛到成员端口link-3作为转发出接口，并基于link-3预留带宽。转发层面，对于TE1新出口link-3，将link-3与SG1形成保护关系。
在实例二的基础上，可选地，之前故障的成员接口链路恢复，或者成员链路添加或减少，或者LSP删除或增加，根据选择策略，TE隧道重新计算新的成员接口，在新的成员接口上预留带宽，并将其与SG口形成接口保护关系。在成员链路1链路故障恢复的情况下，还可以包括再次根据成员接口选择策略，TE1重新使用成员端口链路1作为转发出接口，并基于链路1预留带宽。转发层面，再次将链路1与SG1形成保护关系。
为了尽量在SG TE-TRUNK链路生存期内始终保持均衡负载，从而在故障时更有效地保护LSP，需要在下面情况发生时，对TE-TRUNK链路当前承载的LSP，判断是否需要调整其成员链路位置，从而进行LSP带宽调整：成员链路故障消失；成员链路添加或减少；TE-LSP隧道的删除或增加等。
本发明的目的是提供一种基于捆绑接口的隧道带宽预留的方法。为了保证隧道带宽，提高接口带宽利用率。链路正常情况下固定选择某个成员口进行带宽预留与流量转发，当链路异常时，通过快速切换到捆绑接口其他成员链路进行分担。走相同捆绑接口的不同隧道，选择不同成员链路的情况下，某成员接口故障时，不同隧道间相互不影响。仅对隧道的带宽预留和流量转发进行改进优化，对原有协议层面的协议交互收发包无影响。隧道能够根据带宽约束条件以及捆绑成员链路负载合理分配资源给隧道，同时可以在成员链路故障的时候能够为业务隧道重新选择成员链路，在这种机制下，故障链路的LSP不会重新计算，达到LSP的快速收敛，对业务流量的影响较小。
显然，本领域的技术人员应该明白，上述的本发明的各模块或各步骤可以用通用的计算装置来实现，它们可以集中在单个的计算装置上，或者分布在多个计算装置所组成的网络上，可选地，它们可以用计算装置可执行的程序代码来实现，从而，可以将它们存储在存储装置中由计算装置来执行，并且在某些情况下，可以以不同于此处的顺序执行所示出或描述的步骤，或者将它们分别制作成各个集成电路模块，或者将 它们中的多个模块或步骤制作成单个集成电路模块来实现。这样，本发明不限制于任何特定的硬件和软件结合。
以上所述仅为本发明的优选实施例而已，并不用于限制本发明，对于本领域的技术人员来说，本发明可以有各种更改和变化。凡在本发明的精神和原则之内，所作的任何修改、等同替换、改进等，均应包含在本发明的保护范围之内。
工业实用性
如上所述，本发明实施例提供的一种基于捆绑接口的隧道带宽预留的方法及装置具有以下有益效果：能够保证隧道带宽，提高接口带宽利用率。而且，仅对隧道的带宽预留和流量转发进行改进优化，对原有协议层面的协议交互收发包无影响。此外，隧道能够根据带宽约束条件以及捆绑成员链路负载合理分配资源给隧道，同时可以在成员链路故障的时候能够为业务隧道重新选择成员链路，在这种机制下，故障链路的LSP不会重新计算，达到LSP的快速收敛，对业务流量的影响较小。

Claims (10)

  1. 一种基于捆绑接口的隧道带宽预留的方法，包括：
    根据捆绑接口的成员链路的预设选择策略，选择所述捆绑接口的第一成员链路作为业务隧道的转发表项出接口；
    形成所述第一成员链路与所述捆绑接口中除所述第一成员链路之外的其他有效成员链路之间的保护关系。
  2. 根据权利要求1所述的方法，其中，所述方法还包括：
    在所述第一成员链路上对所述业务隧道进行带宽预留。
  3. 根据权利要求2所述的方法，其中，在所述第一成员链路出现故障的情况下，所述方法还包括：
    将所述业务隧道的流量由所述第一成员链路切换到所述其他有效成员链路，其中，所述其他有效成员链路通过普通负荷分担的方式分担所述业务隧道的流量。
  4. 根据权利要求3所述的方法，其中，将所述业务隧道的流量由所述第一成员链路切换到所述其他有效成员链路之后，所述方法还包括：
    根据所述预设选择策略，选择第二成员链路作为所述业务隧道的转发表项出接口；
    形成所述第二成员链路与所述捆绑接口中除所述第二成员链路之外的其他有效成员链路之间的保护关系；
    在所述第二成员链路上对所述业务隧道进行带宽预留。
  5. 根据权利要求1至4中任一项所述的方法，其中，所述预设选择策略包括以下至少之一：
    在多条成员链路的未预留带宽满足所述业务隧道的带宽需求的情况下，将所述多条成员链路中未预留带宽最小的成员链路作为被选择链路；
    在多条成员链路的未预留带宽满足所述业务隧道的带宽需求的情况下，将所述多条成员链路中索引最小的成员链路作为被选择链路；
    在所述捆绑接口中所有成员链路的未预留带宽均不满足所述业务隧道的带宽需求的情况下，拆除优先级低于所述业务隧道的一个或多个其他业务隧道。
  6. 一种基于捆绑接口的隧道带宽预留的装置，包括：
    第一选择模块，设置为根据捆绑接口的成员链路的预设选择策略，选择所述捆绑接口的第一成员链路作为业务隧道的转发表项出接口；
    第一形成模块，设置为形成所述第一成员链路与所述捆绑接口中除所述第一成员链路之外的其他有效成员链路之间的保护关系。
  7. 根据权利要求6所述的装置，其中，所述装置还包括：
    第一预留模块，设置为在所述第一成员链路上对所述业务隧道进行带宽预留。
  8. 根据权利要求7所述的装置，其中，所述装置还包括：
    切换模块，设置为在所述第一成员链路出现故障的情况下，将所述业务隧道的流量由所述第一成员链路切换到所述其他有效成员链路，其中，所述其他有效成员链路通过普通负荷分担的方式分担所述业务隧道的流量。
  9. 根据权利要求8所述的装置，其中，所述装置还包括：
    第二选择模块，设置为根据所述预设选择策略，选择第二成员链路作为所述业务隧道的转发表项出接口；
    第二形成模块，设置为形成所述第二成员链路与所述捆绑接口中除所述第二成员链路之外的其他有效成员链路之间的保护关系；
    第二预留模块，设置为在所述第二成员链路上对所述业务隧道进行带宽预留。
  10. 根据权利要求6至9中任一项所述的装置，其中，所述预设选择策略包括以下至少之一：
    在多条成员链路的未预留带宽满足所述业务隧道的带宽需求的情况下，将所述多条成员链路中未预留带宽最小的成员链路作为被选择链路；
    在多条成员链路的未预留带宽满足所述业务隧道的带宽需求的情况下，将所述多条成员链路中索引最小的成员链路作为被选择链路；
    在所述捆绑接口中所有成员链路的未预留带宽均不满足所述业务隧道的带宽需求的情况下，拆除优先级低于所述业务隧道的一个或多个其他业务隧道。
PCT/CN2015/077536 2014-12-25 2015-04-27 一种基于捆绑接口的隧道带宽预留的方法及装置 WO2016101469A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201410822029.8 2014-12-25
CN201410822029.8A CN105791108B (zh) 2014-12-25 2014-12-25 一种基于捆绑接口的隧道带宽预留的方法及装置

Publications (1)

Publication Number Publication Date
WO2016101469A1 true WO2016101469A1 (zh) 2016-06-30

Family

ID=56149072

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2015/077536 WO2016101469A1 (zh) 2014-12-25 2015-04-27 一种基于捆绑接口的隧道带宽预留的方法及装置

Country Status (2)

Country Link
CN (1) CN105791108B (zh)
WO (1) WO2016101469A1 (zh)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112953742A (zh) * 2019-12-10 2021-06-11 中盈优创资讯科技有限公司 网络设备捆绑端口的故障定位方法及装置
CN115022248A (zh) * 2022-05-25 2022-09-06 烽火通信科技股份有限公司 一种HQoS业务接入方法及装置

Families Citing this family (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109309530B (zh) * 2017-07-28 2023-08-01 中兴通讯股份有限公司 一种数据传输方法和装置
CN110417651B (zh) 2018-04-28 2021-07-16 华为技术有限公司 一种隧道建立方法、装置及系统
CN110611577A (zh) * 2018-06-14 2019-12-24 中兴通讯股份有限公司 业务快速切换方法、切换装置、网络设备和存储介质
CN110944357B (zh) * 2018-09-25 2023-08-29 中兴通讯股份有限公司 一种网络切片的方法及装置
CN111031516B (zh) * 2018-10-10 2024-04-09 上汽通用汽车有限公司 用于为车载通信系统提供安全防护的方法和装置
CN112788637A (zh) * 2020-06-29 2021-05-11 中兴通讯股份有限公司 隧道保护方法、边缘路由器和计算机可读介质

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1866923A (zh) * 2006-02-24 2006-11-22 华为技术有限公司 一种实现捆绑接口边到边伪线仿真服务的方法及系统
CN101674224A (zh) * 2008-09-08 2010-03-17 华为技术有限公司 生成转发表项信息的方法、标签交换路由器及系统
CN101834759A (zh) * 2010-04-21 2010-09-15 中兴通讯股份有限公司 捆绑链路的检测方法及分布式设备
CN102271048A (zh) * 2010-06-02 2011-12-07 中兴通讯股份有限公司 聚合链路中的业务保护方法及装置
CN102761480A (zh) * 2012-06-28 2012-10-31 中兴通讯股份有限公司 一种资源分配方法及装置

Family Cites Families (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101197773B (zh) * 2007-10-25 2010-12-15 华为技术有限公司 基于多链路的报文压缩发送和接收解压缩方法及系统
US8908662B2 (en) * 2009-02-11 2014-12-09 Futurewei Technologies, Inc. Apparatus and method of flow movement for network-based mobility management protocol
CN101841487A (zh) * 2010-05-24 2010-09-22 中兴通讯股份有限公司 聚合链路服务流的配置方法及包交换装置
US20130036032A1 (en) * 2011-08-04 2013-02-07 Yigang Cai Service plan negotiations with end users for policy and charging control (pcc)

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1866923A (zh) * 2006-02-24 2006-11-22 华为技术有限公司 一种实现捆绑接口边到边伪线仿真服务的方法及系统
CN101674224A (zh) * 2008-09-08 2010-03-17 华为技术有限公司 生成转发表项信息的方法、标签交换路由器及系统
CN101834759A (zh) * 2010-04-21 2010-09-15 中兴通讯股份有限公司 捆绑链路的检测方法及分布式设备
CN102271048A (zh) * 2010-06-02 2011-12-07 中兴通讯股份有限公司 聚合链路中的业务保护方法及装置
CN102761480A (zh) * 2012-06-28 2012-10-31 中兴通讯股份有限公司 一种资源分配方法及装置

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112953742A (zh) * 2019-12-10 2021-06-11 中盈优创资讯科技有限公司 网络设备捆绑端口的故障定位方法及装置
CN112953742B (zh) * 2019-12-10 2023-05-12 中盈优创资讯科技有限公司 网络设备捆绑端口的故障定位方法及装置
CN115022248A (zh) * 2022-05-25 2022-09-06 烽火通信科技股份有限公司 一种HQoS业务接入方法及装置
CN115022248B (zh) * 2022-05-25 2023-11-03 烽火通信科技股份有限公司 一种HQoS业务接入方法及装置

Also Published As

Publication number Publication date
CN105791108B (zh) 2019-07-12
CN105791108A (zh) 2016-07-20

Similar Documents

Publication Publication Date Title
WO2016101469A1 (zh) 一种基于捆绑接口的隧道带宽预留的方法及装置
US9253097B1 (en) Selective label switched path re-routing
US8462783B2 (en) Method and a node device for transferring a message based on traffic engineering tunnels
US10313234B2 (en) RSVP make-before-break label reuse
US8451846B1 (en) LSP hierarchy for MPLS networks
EP2678982B1 (en) System and method for advertising a composite link in interior gateway protocol and/or interior gateway protocol-traffic engineering
CN112054969B (zh) 一种实现报文镜像的方法及装置
CN101420384A (zh) 一种实现ecmp域内自愈保护的方法
US7596088B2 (en) Route selection with bandwidth sharing optimization over rings
US20070140233A1 (en) Resource sharing among network
CN103441930B (zh) 一种mpls te分组转发与管理方法及装置
EP3493492B1 (en) Rsvp make-before-break label reuse
WO2012068911A1 (zh) 一种保护带宽资源的预留方法及系统
WO2012103729A1 (zh) 隧道配置方法及装置
Petersson MPLS based recovery mechanisms
US9876675B1 (en) Packet transport shared mesh protection
JP2002094555A (ja) ラベルスイッチングパスの優先制御方法およびこれを用いた通信システム
Chaieb et al. A new pre-emption policy for mpls-te networks
JP2004260696A (ja) フレーム転送ネットワーク
WO2016177186A1 (zh) 带宽保护的方法及装置
Hassan et al. Priority-based allocation of network resources in multi-class MPLS networks
IL192993A (en) Routing selection with bandwidth sharing optimization on rings

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: 15871545

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 15871545

Country of ref document: EP

Kind code of ref document: A1