WO2017028564A1 - 一种交换网组播路由的实现方法和装置、计算机存储介质 - Google Patents

一种交换网组播路由的实现方法和装置、计算机存储介质 Download PDF

Info

Publication number
WO2017028564A1
WO2017028564A1 PCT/CN2016/082149 CN2016082149W WO2017028564A1 WO 2017028564 A1 WO2017028564 A1 WO 2017028564A1 CN 2016082149 W CN2016082149 W CN 2016082149W WO 2017028564 A1 WO2017028564 A1 WO 2017028564A1
Authority
WO
WIPO (PCT)
Prior art keywords
forwarding
routing
link
algorithm
chips
Prior art date
Application number
PCT/CN2016/082149
Other languages
English (en)
French (fr)
Inventor
汪为汉
Original Assignee
深圳市中兴微电子技术有限公司
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 深圳市中兴微电子技术有限公司 filed Critical 深圳市中兴微电子技术有限公司
Publication of WO2017028564A1 publication Critical patent/WO2017028564A1/zh

Links

Images

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L45/00—Routing or path finding of packets in data switching networks
    • H04L45/16—Multipoint routing
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L43/00—Arrangements for monitoring or testing data switching networks
    • H04L43/08—Monitoring or testing based on specific metrics, e.g. QoS, energy consumption or environmental parameters
    • H04L43/0823—Errors, e.g. transmission errors
    • H04L43/0829—Packet loss
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L47/00—Traffic control in data switching networks
    • H04L47/70—Admission control; Resource allocation
    • H04L47/80—Actions related to the user profile or the type of traffic
    • H04L47/806—Broadcast or multicast traffic
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L49/00—Packet switching elements
    • H04L49/10—Packet switching elements characterised by the switching fabric construction
    • H04L49/109—Integrated on microchip, e.g. switch-on-chip

Definitions

  • the invention relates to a data routing and switching technology, in particular to a method and a device for implementing a multicast routing of a switching network, and a computer storage medium.
  • multicast can substantially reduce the operating cost and bandwidth requirements of the entire network, reduce the load on the server and the network, and improve the service quality of the network.
  • the multicast routing table structure configured by the upper layer software is difficult to respond to the system or the switch chip in real time.
  • the embodiment of the present invention is to provide a method and a device for implementing a multicast routing of a switching network, and a computer storage medium, which can effectively avoid the loss of multicast cells to a certain extent, thereby improving the exchange of multicast cells. effectiveness.
  • An embodiment of the present invention provides a method for implementing a multicast routing of a switching network, where the method includes:
  • a final link is selected for routing forwarding.
  • the acquiring a set of target forwarding chips that meet the forwarding condition is:
  • Matching criteria are used between adjacent two-stage switching chips to find a conditional expression A class of target forwarding chip sets.
  • selecting a final link for routing and forwarding includes:
  • a polling (RR, Round-Robin) algorithm randomly selects one link for routing and forwarding in an aggregate link having the same forwarding capability.
  • the RR algorithm includes a Weighted Round Robin (WRR) algorithm, a Defict Round-Robin (DRR) algorithm, and an Uruguur-based Round-Robin (URR) algorithm.
  • WRR Weighted Round Robin
  • DRR Defict Round-Robin
  • URR Uruguur-based Round-Robin
  • the matched position is between any adjacent two-stage switching chips.
  • An embodiment of the present invention further provides an apparatus for implementing a multicast routing of a switching network, where the apparatus includes:
  • Obtaining a module configured to acquire a set of target forwarding chips that meet forwarding conditions
  • a link establishing module configured to: establish an optional set link according to the obtained set of target forwarding chips that meet the forwarding condition;
  • the routing module is configured to select a final link for routing and forwarding in the optional aggregation link.
  • the acquiring module is configured to find a conditional condition by using a matching criterion between adjacent two-stage switching chips.
  • the routing module is configured to randomly select a link for routing and forwarding in the aggregate link with the same forwarding capability by using the polling RR algorithm in the optional aggregation link.
  • the RR algorithm includes a weighted polling WRR algorithm, a difference polling DRR algorithm, and an emergency polling URR algorithm.
  • the matched position is between any adjacent two-stage switching chips.
  • the obtaining module, the link establishing module, and the routing module may perform a processing by using a central processing unit (CPU), a digital signal processor (DSP), or a programmable logic.
  • CPU central processing unit
  • DSP digital signal processor
  • Array FPGA, Field-Programmable Gate Array
  • the embodiment of the invention further provides a computer storage medium, wherein computer executable instructions are stored, and the computer executable instructions configure an implementation method for performing the above-mentioned switching network multicast routing.
  • the method and device for implementing the multicast routing of the switching network provided by the embodiment of the present invention, by selecting the exact matching of the next-level routing, adopting the method of multicast routing replication to perform route differentiation, so that the replication of the multicast routing can be performed to a certain extent.
  • the secure and efficient exchange in the network effectively solves the problem of multicast cell loss in the distributed network, thereby improving the switching efficiency and user experience of the multicast cell.
  • FIG. 1 is a schematic diagram of a topology structure of an existing three-level CLOS distributed networking
  • FIG. 2 is a schematic diagram of an implementation process of an existing multicast route
  • FIG. 3 is a diagram showing an example of packet loss of a multicast cell
  • FIG. 4 is a schematic flowchart of an implementation process of a multicast routing method according to the present invention.
  • FIG. 5 is a schematic diagram of a route forwarding process using a polling exact matching method according to the present invention.
  • FIG. 6 is a schematic structural diagram of a device for multicast routing in the present invention.
  • the distributed networking is taken as an example to further explain the implementation process of the existing distributed network topology and multicast routing.
  • the three-level CLOS distributed networking topology popular in the industry is taken as an example for description.
  • the existing three-level CLOS distributed networking topology is as shown in FIG. 1.
  • the three-level CLOS distributed networking includes: n switching access chips S i (0 ⁇ i ⁇ n), n first-level switching Chip A j (0 ⁇ j ⁇ n ), m second-stage central switching chips B k (0 ⁇ k ⁇ m), n third-stage switching chips C g (0 ⁇ g ⁇ n), n switching connections Chip S h (0 ⁇ h ⁇ n).
  • the three-level CLOS distributed networking can transmit not only unicast cells but also multicast cells; in FIG. 1, To represent unicast cells, To represent multicast cells.
  • Any switch chip of any level to all switch chips of the next level or the previous level has one and only one connection; then, if there are two or more routing paths, the exchange of the same two adjacent levels The chip, the two routing paths are blocked.
  • the routing path from input to output is not unique. For example, different paths can be selected according to the selection of the second-level central switching chip B k (0 ⁇ k ⁇ m); and the second-level central switching chip B k The more you choose, the more paths you have and the less likely you are to block.
  • FIG. 2 is a schematic diagram of an implementation process of an existing multicast routing. As shown in FIG. 2, it is assumed that only a multicast cell is transmitted in the three-level CLOS distributed networking, and is used in FIG. To represent multicast cells.
  • the switch access chip S i (0 ⁇ i ⁇ n) and the n switch take-off chips S h (0 ⁇ h ⁇ n) are combined into four racks.
  • the software system is mainly used to obtain the system topology, and the secondary replication is used to complete the multicast replication in the switching network, and then the copied multicast cells are sent to the destination ports to complete one-to-many conversion.
  • the process is as follows:
  • B 1 chip A type C chip is selected from the set ⁇ C 1 , C 2 ... C k ⁇ for routing by extensive polling, and then copied by the chip, and routed to each destination chip ⁇ S 1 , S 2 ...S k ⁇ .
  • the method for ⁇ C 1, C 2 ... C k ⁇ forwarding chip routing capability is not distinguished, and in the actual networking environment, ⁇ C 1, C 2 ... C k ⁇ in The routing capability of each chip to the rack 0 ⁇ S 1 , S 2 ... S k ⁇ chip must also be strictly equivalent.
  • the multicast replication software configuration achieves a structural lag in time, resulting in a topology change caused by the inability of the switch chip to sense the link state in real time.
  • Some edge switch chips C k or the center switch chip B k are in a period of time. It is still regarded as a copy forwarding path, which eventually leads to continuous packet loss of multicast cells.
  • the embodiment of the present invention avoids packet loss of multicast cells by means of multicast route replication.
  • a new group proposed by the embodiment of the present invention is further described in detail by taking a multicast copy of the B 1 chip to the k-shared multicast cells in the rack 0 ⁇ S 1 , S 2 ... S k ⁇ chip.
  • the specific process of broadcasting routes is further described in detail by taking a multicast copy of the B 1 chip to the k-shared multicast cells in the rack 0 ⁇ S 1 , S 2 ... S k ⁇ chip.
  • FIG. 4 is a schematic flowchart of an implementation of a method for multicast routing according to an embodiment of the present invention.
  • the embodiment of the present invention is implemented in the network environment shown in FIG. 2, as shown in FIG. 4, the method for multicast routing in the embodiment of the present invention is shown in FIG.
  • the implementation process includes the following steps:
  • Step 401 Acquire a set of target forwarding chips that meet forwarding conditions.
  • the forwarding condition may be capable of routing differentiation, and the routing capability of each chip in the forwarding process to the target chip is strictly equivalent.
  • Matching criteria are used between adjacent two-stage switching chips to find a conditional expression A class of target forwarding chip sets.
  • Step 402 Establish an optional set link according to the obtained set of target forwarding chips that meet the forwarding condition.
  • the establishing an optional set link is: establishing a routing link between the switching chip to be forwarded and the target forwarding chip set to form a selectable aggregate link; how to establish a routing link belongs to the existing Technology will not be described here.
  • Step 403 In the optional aggregation link, select a final link for route forwarding.
  • the selection may be that, by using an RR algorithm, in a set link having the same forwarding capability, one link is randomly selected for routing and forwarding.
  • the RR algorithm includes a WRR algorithm, a DRR algorithm, and a URR algorithm.
  • the embodiment of the present invention takes the B 1 chip multicast replication to the k-shared multicast cells in the rack 0 ⁇ S 1 , S 2 ... S k ⁇ chip as an example, and adopts a route forwarding process in which the polling exact matching method is adopted. As shown in FIG. 5, the specific includes:
  • each multicast routing process B 1 C i after completion of the ⁇ S i ... S j ⁇ k parts of replication, a selection of C i on the edges of the chip, extensive polling change the original selection method, Using a precise matching method, the class C edge-switching chip that cannot satisfy the forwarding condition is eliminated, and a class of target forwarding chip sets is obtained. After obtaining a set of target forwarding chips, in the aggregate link of the same forwarding capability, a final link is selected by the RR algorithm for routing and forwarding, thereby completing the routing process from B 1 to C i , and then the C i chip Complete the copying process to the ⁇ S i ... S j ⁇ exchange access chip.
  • the exact matching manner refers to: finding a conditional expression in the set of C i n (1 ⁇ n ⁇ k) sets of C-type chips.
  • Another embodiment of the present invention extends the above method to the multicast routing process of the primary edge switching chip A to the tertiary edge switching chip C, and completes the multicast routing and copying process.
  • the multicast replication and routing process of the entire switching network is designed to complete the one-to-many routing process of the switching access chip S i (0 ⁇ i ⁇ n) to the switching chip S h (0 ⁇ h ⁇ n).
  • the transmission of the multicast cell needs to go through the four routing routing process of the intermediate switching layer.
  • the specific implementation process is as follows:
  • the first level of routing the switching access chip S i sends the cell to the primary edge switching chip A i , and selects one of the primary edge switching chips A i to exchange;
  • multicast replication is not performed, and the multicast cell is processed according to the unicast cell routing table.
  • Second-level routing This level of routing can complete the routing and routing process of the cell-level edge switching chip A i to the central switching chip B i .
  • the exact matching method is adopted to eliminate the B that cannot satisfy the forwarding condition.
  • Class-center switching chip get target forwarding chip set After obtaining the target forwarding chip set, in the aggregate link of the same forwarding capability, the RR algorithm is used to select a final link for transmission, and the routing process from A 1 to B i is completed.
  • the selected central switching chip B i can complete one to many routing processes, and can also meet the requirements of route replication.
  • This level of routing can complete the routing of the tertiary edge switching chip C i and complete the primary replication process. Taking the central switching chip B 1 as an example, in each multicast routing process, B 1 is completed by C i to the k copy of ⁇ S i ...
  • Fourth-level routing After the multicast cell reaches the tertiary edge chip C i , it only needs to complete the replication and routing process in the rack.
  • the original routing policy can be used, the system configuration forwarding table is adopted, or the system unicast routing path is used to complete the last copy forwarding routing process of the multicast cell.
  • the embodiment of the present invention further provides an apparatus for implementing a multicast routing of a switching network.
  • the apparatus includes an obtaining module 61, a link establishing module 62, and a routing module 63.
  • the obtaining module 61 is configured to acquire a set of target forwarding chips that meet the forwarding condition
  • the link establishing module 62 is configured to establish an optional set link according to the acquired set of target forwarding chips that meet the forwarding condition.
  • the routing module 63 is configured to select a final link for routing forwarding in the optional aggregation link.
  • the obtaining module 61 is configured to find a conditional condition by using a matching criterion between adjacent two-stage switching chips.
  • the routing module 63 is configured to randomly select one of the set links with the same forwarding capability by using the polling RR algorithm in the optional aggregation link.
  • the link performs route forwarding.
  • the RR algorithm includes a weighted round-robin WRR algorithm, a differential polling DRR algorithm, and an emergency polling URR algorithm.
  • the matching location is between any two adjacent switching chips.
  • the obtaining module 61, the link establishing module 62, and the routing module 63 can be a Central Processing Unit (CPU), a Micro Processor Unit (MPU), a Digital Signal Processor (DSP), or a Field Programmable Gate Array (FPGA) located in the routing device. , Field Programmable Gate Array) and other implementations.
  • CPU Central Processing Unit
  • MPU Micro Processor Unit
  • DSP Digital Signal Processor
  • FPGA Field Programmable Gate Array
  • the routing of the multicast routing is performed by using the method of the multicast routing, so that the replication of the multicast routing can adapt to the change of the network topology to adjust the routing policy in time;
  • the multicast cell loss can be effectively avoided to a certain extent, thereby ensuring the secure and efficient exchange of the multicast cell in the network, and effectively solving the distributed problem.
  • the problem of loss of multicast cells in the network improves the switching efficiency and user experience of multicast cells.
  • the embodiment of the invention further provides a computer storage medium, wherein computer executable instructions are stored, and the computer executable instructions configure an implementation method for performing the above-mentioned switching network multicast routing.
  • the implementation scheme of the multicast routing of the switching network performs routing matching by using the method of multicast routing replication, so that the replication of the multicast routing can be adapted to a certain extent.
  • the network topology changes, so that the routing strategy is adjusted in time; thus, the path of the existing multicast routing can be optimized to avoid the loss of multicast cells to a certain extent, thereby ensuring that the multicast cells are in the network. Safe and efficient switching effectively solves the problem of multicast cell loss in distributed networks, thereby improving the switching efficiency and user experience of multicast cells.

Landscapes

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

Abstract

本发明实施例公开了一种交换网组播路由的实现方法，包括：获取满足转发条件的一类目标转发芯片集合；根据获取的所述满足转发条件的一类目标转发芯片集合，建立可选的集合链路；在所述可选的集合链路中，选择一条最终链路进行路由转发。本发明还同时公开了一种交换网组播路由的实现装置、计算机存储介质。

Description

一种交换网组播路由的实现方法和装置、计算机存储介质 技术领域
本发明涉及数据路由交换技术，尤其涉及一种交换网组播路由的实现方法和装置、计算机存储介质。
背景技术
随着网络数据的急剧增长，传统的以单级路由交换芯片组网而成的交换网络，都会浪费大量的带宽资源，难以满足交换数据容量的快速增长。为解决该技术问题，一种由交换芯片构成的分布式集群网络成为新的发展方向，用以实现更大容量的数据交换，因此，如何在更复杂的组网结构中寻找有效的组播路由路径具有重要意义。
组播虽然可以从本质上减少整个网络的运营代价和带宽需求，降低服务器和网络的负载，提高网络的服务质量。但是，现有的组播路由交换策略中，当系统的拓扑结构发生变化，或者交换芯片间的链路状态发生变化时，由上层软件配置的组播路由表结构难以实时响应系统或交换芯片间的拓扑变化、更新配置信息，因此会导致组播信元的持续丢包，这样，不仅降低了组播信元在网络中安全、高效的交换效率，还给用户的正常使用带来了一定的影响。
发明内容
有鉴于此，本发明实施例期望提供一种交换网组播路由的实现方法和装置、计算机存储介质，能够在一定程度上有效地避免组播信元的丢失，从而提高组播信元的交换效率。
为达到上述目的，本发明实施例的技术方案是这样实现的：
本发明实施例提供一种交换网组播路由的实现方法，所述方法包括：
获取满足转发条件的一类目标转发芯片集合；
根据获取的所述满足转发条件的一类目标转发芯片集合，建立可选的集合链路；
在所述可选的集合链路中，选择一条最终链路进行路由转发。
上述方案中，所述获取满足转发条件的一类目标转发芯片集合为：
在相邻的两级交换芯片之间，采用匹配基准找到满足条件式
Figure PCTCN2016082149-appb-000001
的一类目标转发芯片集合。
上述方案中，所述在所述可选的集合链路中，选择一条最终链路进行路由转发包括：
在所述可选的集合链路中，通过轮询(RR，Round-Robin)算法，在具有相同转发能力的集合链路中，随机选择一条链路进行路由转发。
上述方案中，所述RR算法包括加权轮询(WRR，Weight Round-Robin)算法、差额轮询(DRR，Defict Round-Robin)算法、紧急轮询(URR，Urgency-based Round-Robin)算法。
上述方案中，所述匹配的位置处于任意相邻的两级交换芯片之间。
本发明实施例还提供一种交换网组播路由的实现装置，所述装置包括：
获取模块，配置为获取满足转发条件的一类目标转发芯片集合；
链路建立模块，配置为根据获取的所述满足转发条件的一类目标转发芯片集合，建立可选的集合链路；
路由选择模块，配置为在所述可选的集合链路中，选择一条最终链路进行路由转发。
上述方案中，所述获取模块，配置为在相邻的两级交换芯片之间，采用匹配基准找到满足条件式
Figure PCTCN2016082149-appb-000002
的一类目标转发芯片集合。
上述方案中，所述路由选择模块，配置为在所述可选的集合链路中，通过轮询RR算法，在具有相同转发能力的集合链路中，随机选择一条链路进行路由转发。
上述方案中，所述RR算法包括加权轮询WRR算法、差额轮询DRR算法、紧急轮询URR算法。
上述方案中，所述匹配的位置处于任意相邻的两级交换芯片之间。
所述获取模块、所述链路建立模块、所述路由选择模块在执行处理时，可以采用中央处理器(CPU，Central Processing Unit)、数字信号处理器(DSP，Digital Singnal Processor)或可编程逻辑阵列(FPGA，Field－Programmable Gate Array)实现。
本发明实施例还提供一种计算机存储介质，其中存储有计算机可执行指令，该计算机可执行指令配置执行上述交换网组播路由的实现方法。
本发明实施例所提供的交换网组播路由的实现方法和装置，通过对下一级路由的精确匹配选择，采用组播路由复制的方式进行路由区分，使得组播路由的复制能够在一定程度上适应网络拓扑的变化，从而及时调整路由策略；如此，就能通过优化现有组播路由的寻径过程，在一定程度上有效地避免组播信元的丢失，从而保证组播信元在网络中安全、高效的交换，有效地解决了分布式网络中组播信元丢失的问题，进而提高了组播信元的交换效率及用户体验。
附图说明
图1为现有三级CLOS分布式组网拓扑结构示意图；
图2为现有组播路由的实现过程示意图；
图3为组播信元丢包的示例图；
图4为本发明中组播路由的方法的实现流程示意图；
图5为本发明中采用轮询精确匹配方式的路由转发流程示意图；
图6为本发明中组播路由的装置的组成结构示意图。
具体实施方式
为了能更清楚地说明本发明实施例所解决的技术问题，先以分布式组网为例，进一步说明现有分布式网络拓扑结构和组播路由的实现过程。这里，以业界流行的三级CLOS分布式组网拓扑结构为例进行说明。
现有三级CLOS分布式组网拓扑结构如图1所示，所述三级CLOS分布式组网包括：n个交换接入芯片Si(0≤i≤n)、n个第一级交换芯片Aj(0≤j≤n)、m个第二级中心交换芯片Bk(0≤k≤m)、n个第三级交换芯片Cg(0≤g≤n)、n个交换接出芯片Sh(0≤h≤n)。其中，所述三级CLOS分布式组网中不仅可以传输单播信元，还可以传输多播信元；图1中，用
Figure PCTCN2016082149-appb-000003
来表示单播信元，用
Figure PCTCN2016082149-appb-000004
来表示多播信元。
任意一级的交换芯片到下一级或者前一级的所有交换芯片，都有且只有一个连接；那么，如果有两条或者两条以上的路由路径，具有相同的两个相邻级的交换芯片，则这两条路由路径阻塞。如此，从输入到输出的路由路径不是唯一的，例如，根据第二级中心交换芯片Bk(0≤k≤m)的选择不同，可以选择不同的路径；且第二级中心交换芯片Bk越多，可选择的路径就越多，阻塞的概率就越小。
图2为现有组播路由的实现过程示意图，如图2所示，假设该三级CLOS分布式组网中只传输多播信元，图2中用
Figure PCTCN2016082149-appb-000005
来表示多播信元。图2中，将该三级路由交换系统中的n个第一级交换芯片Aj(0≤j≤n)和n个第三级交换芯片Cg(0≤g≤n)，与n个交换接入芯片Si(0≤i≤n)和n个交换接出芯片Sh(0≤h≤n)组合为四个机架。
其中，第三级交换芯片Cg(0≤g≤n)中的部分集合芯片{C1,C2...Ck}与交换接出芯片{S1,S2...Sk}组合为机架0，第三级交换芯片Cg(0≤g≤n)中剩余的部 分集合芯片{Ck+x...Cn}与交换接出芯片{Sk+x...Sn}组合为机架1，第一级交换芯片Aj(0≤j≤n)中的部分集合芯片{A1,A2...Ak}与交换接入芯片{S1,S2...Sk}组合为机架2，第一级交换芯片Aj(0≤j≤n)中剩余的部分集合芯片{Ak+x...An}与交换接入芯片{Sk+x...Sn}组合为机架3。
这里，主要利用软件系统获取系统拓扑，采用二级复制方式完成在交换网中的组播复制，然后再将复制后的组播信元发往各个目的端口，完成一对多的转换，具体实现过程如下：
首先，组播信元通过第一级交换芯片Aj，传输到第二级中心交换芯片Bk；其次，第二级中心交换芯片Bk对输入的信元进行复制，路由寻径后传输到第三级交换芯片Cg；最后，由第三级交换芯片Cg完成组播信元的再次复制、路由寻径后传输到最终的目的端口。
这里，以B1芯片组播复制到机架0{S1,S2...Sk}芯片中k份组播信元为例，现有组播路由的具体实现流程为：B1芯片采用粗放式轮询方式从集合{C1,C2...Ck}中选出一片C类芯片进行路由，再由该芯片进行复制，路由发往各目的端芯片{S1,S2...Sk}。可以发现，该方法对于{C1,C2...Ck}中的芯片转发能力不进行路由区分，且在实际的组网环境中，{C1,C2...Ck}中各芯片对机架0{S1,S2...Sk}芯片的路由能力也并一定是严格对等的。
从上述实际应用的实现过程可以看出，现有的组播路由交换策略中，当三级CLOS分布式网络系统的拓扑结构发生变化，或者交换芯片间的链路状态发生变化时，由上层软件配置的组播路由表结构难以实时响应系统或交换芯片间的拓扑变化、更新配置信息，这就会导致组播信元的持续丢包。
通常，组播信元丢包的示例如图3所示，主要包括以下两种场景：
(1)当三级CLOS分布式网络系统的拓扑结构或交换芯片间的链路状 态发生变化时，导致交换网络中某个边缘交换芯片Ck与交换接出芯片Sk非全路由可达时，中心交换芯片Bk就无法通过此边缘交换芯片Ck完成对交换接出芯片Sk的全复制；
(2)当中心交换芯片Bk到边缘交换芯片Ck的路径上的链路无效时，该中心交换芯片Bk则无法完成组播信元到边缘交换芯片Ck的复制和路由过程。
因此，组播复制软件配置实现结构时间上的滞后性，导致了交换芯片之间不能实时地感知链路状态而引起的拓扑变化，某些边缘交换芯片Ck或者中心交换芯片Bk在一段时间内仍被当作复制转发路径，最终导致组播信元的持续丢包。
为解决上述技术问题，本发明实施例正是通过组播路由复制的方式来避免组播信元的丢包。
本发明实施例以B1芯片组播复制到机架0{S1,S2...Sk}芯片中k份组播信元为例，进一步详细说明本发明实施例提出的新的组播路由的具体流程。
为了能够更加详尽地了解本发明实施例的特点与技术内容，下面结合附图对本发明实施例的实现进行详细阐述，所附附图仅供参考说明之用，并非用来限定本发明。
图4为本发明实施例中组播路由的方法的实现流程示意图，本发明实施例在基于图2所示的网络环境下实现，如图4所示，本发明实施例中组播路由的方法的实现流程包括以下步骤：
步骤401：获取满足转发条件的一类目标转发芯片集合；
这里，所述转发条件可以是能够进行路由区分的，且转发过程中的各芯片对目标芯片的路由能力是严格对等的。在相邻的两级交换芯片之间，采用匹配基准找到满足条件式
Figure PCTCN2016082149-appb-000006
的一类目标转发芯片集合。
步骤402：根据获取的所述满足转发条件的一类目标转发芯片集合，建立可选的集合链路；
这里，所述建立可选的集合链路为：在要转发的交换芯片与目标转发芯片集合之间一一建立路由链路，形成可以选择的集合链路；具体如何建立路由链路属于现有技术，在此不再赘述。
步骤403：在所述可选的集合链路中，选择一条最终链路进行路由转发。
这里，所述选择可以是通过RR算法，在具有相同转发能力的集合链路中，随机选择一条链路进行路由转发。其中，所述RR算法包括WRR算法、DRR算法、URR算法。
其中，本发明实施例以B1芯片组播复制到机架0{S1,S2...Sk}芯片中k份组播信元为例，采用轮询精确匹配方式的路由转发流程如图5所示，具体包括：
在每次组播路由过程中，B1经过Ci完成到{Si...Sj}的k份复制，在对Ci边缘芯片的选择上，改变原有粗放式轮询选择方式，采用精确匹配方式，剔除不能满足转发条件的C类边缘交换芯片，得到一类目标转发芯片集合
Figure PCTCN2016082149-appb-000007
得到一类目标转发芯片集合后，在相同转发能力的集合链路中，通过RR算法，选择一条最终链路进行路由转发，进而完成从B1到Ci的路由选择过程，再由Ci芯片完成到{Si...Sj}交换接入芯片的复制流程。
这里，对于B1的目标芯片集合{Si...Sj}匹配过程，所述精确匹配方式是指：在Ci的集合中，找到满足条件式
Figure PCTCN2016082149-appb-000008
的n(1≤n≤k)个C类芯片集合。
本发明的另一个实施例是将上述方法扩展应用到一级边缘交换芯片A到三级边缘交换芯片C的组播路由选择过程之中，完成组播的路由选择和复制过程。
下面结合附图对本发明实施例提出的一种交换网组播路由的实现方法 在整个分布式交换网中的实施过程作进一步的详细描述。
整个交换网络的组播复制及路由过程，旨在完成交换接入芯片Si(0≤i≤n)对交换接出芯片Sh(0≤h≤n)的一次一对多的路由过程，其中，组播信元的传递需要经过中间交换层的四次路由寻径过程，具体实施过程如下：
第一级路由：交换接入芯片Si将信元发往一级边缘交换芯片Ai，选择一片一级边缘交换芯片Ai上交换；
这里，在该第一级路由中，并不进行组播复制，且组播信元按照单播信元路由查表处理。
第二级路由：该级路由可以完成信元一级边缘交换芯片Ai到中心交换芯片Bi的传递路由选择过程，以A1芯片为例，采用精确匹配方式，剔除不能满足转发条件的B类中心交换芯片，得到目标转发芯片集合
Figure PCTCN2016082149-appb-000009
得到目标转发芯片集合后，在相同转发能力的集合链路中，通过RR算法，选择一条最终链路进行发送，完成从A1到Bi的路由选择过程。
这里，对于A1的目标芯片集合{Si...Sj}匹配过程，主要是在Bi的集合中，找到满足条件式
Figure PCTCN2016082149-appb-000010
的n(1≤n≤k)个B类芯片集合；所选的中心交换芯片Bi能够完成一到多的路由过程，且还能满足路由复制的要求。
第三级路由：该级路由可以完成对三级边缘交换芯片Ci的路由选择，同时完成一级复制过程。以中心交换芯片B1为例，在每次组播路由过程，B1经过Ci完成到{Si...Sj}的k份复制，在对Ci边缘芯片的选择上，采用与第二级路由中选择中心交换芯片Bi的相同方法；即：找到满足条件式
Figure PCTCN2016082149-appb-000011
的n(1≤n≤k)个C类芯片集合，得到目标转发芯片集合
Figure PCTCN2016082149-appb-000012
得到目标转发芯片集合后，在相同转发能力的集合链 路中，通过RR算法，选择一条最终链路进行发送，完成从B1到Ci的路由选择过程。
第四级路由：组播信元到达三级边缘芯片Ci后，只需要完成在机架内的复制和路由过程。
这里，可沿用原有路由策略，采用系统配置转发表或者采用系统单播路由路径完成组播信元的最后复制转发路由过程。
为实现上述方法，本发明实施例还提供了一种交换网组播路由的实现装置，如图6所示，该装置包括获取模块61、链路建立模块62和路由选择模块63；其中，
获取模块61，配置为获取满足转发条件的一类目标转发芯片集合；
链路建立模块62，配置为根据获取的所述满足转发条件的一类目标转发芯片集合，建立可选的集合链路；
路由选择模块63，配置为在所述可选的集合链路中，选择一条最终链路进行路由转发。
这里，所述获取模块61，配置为在相邻的两级交换芯片之间，采用匹配基准找到满足条件式
Figure PCTCN2016082149-appb-000013
的一类目标转发芯片集合。
在本发明实施例一实施方式中，所述路由选择模块63，配置为在所述可选的集合链路中，通过轮询RR算法，在具有相同转发能力的集合链路中，随机选择一条链路进行路由转发。
在本发明实施例一实施方式中，所述RR算法包括加权轮询WRR算法、差额轮询DRR算法、紧急轮询URR算法。
在本发明实施例一实施方式中，所述匹配的位置处于任意相邻的两级交换芯片之间。
在实际应用中，所述获取模块61、链路建立模块62和路由选择模块 63均可由位于路由设备中的中央处理器(CPU，Central Processing Unit)、微处理器(MPU，Micro Processor Unit)、数字信号处理器(DSP，Digital Signal Processor)、或现场可编程门阵列(FPGA，Field Programmable Gate Array)等实现。
本发明实施例通过对下一级路由的精确匹配选择，采用组播路由复制的方式进行路由区分，使得组播路由的复制能够在一定程度上适应网络拓扑的变化，从而及时调整路由策略；如此，就能通过优化现有组播路由的寻径过程，在一定程度上有效地避免组播信元的丢失，从而保证组播信元在网络中安全、高效的交换，有效地解决了分布式网络中组播信元丢失的问题，进而提高了组播信元的交换效率及用户体验。
本发明实施例还提供一种计算机存储介质，其中存储有计算机可执行指令，该计算机可执行指令配置执行上述交换网组播路由的实现方法。
以上所述，仅为本发明的较佳实施例而已，并非用于限定本发明的保护范围，凡在本发明的精神和原则之内所作的任何修改、等同替换和改进等，均应包含在本发明的保护范围之内。
工业实用性
本发明实施例所提供的交换网组播路由的实现方案，通过对下一级路由的精确匹配选择，采用组播路由复制的方式进行路由区分，使得组播路由的复制能够在一定程度上适应网络拓扑的变化，从而及时调整路由策略；如此，就能通过优化现有组播路由的寻径过程，在一定程度上有效地避免组播信元的丢失，从而保证组播信元在网络中安全、高效的交换，有效地解决了分布式网络中组播信元丢失的问题，进而提高了组播信元的交换效率及用户体验。

Claims (11)

  1. 一种交换网组播路由的实现方法，所述方法包括：
    获取满足转发条件的一类目标转发芯片集合；
    根据获取的所述满足转发条件的一类目标转发芯片集合，建立可选的集合链路；
    在所述可选的集合链路中，选择一条最终链路进行路由转发。
  2. 根据权利要求1所述的方法，其中，所述获取满足转发条件的一类目标转发芯片集合为：
    在相邻的两级交换芯片之间，采用匹配基准找到满足条件式
    Figure PCTCN2016082149-appb-100001
    的一类目标转发芯片集合。
  3. 根据权利要求1所述的方法，其中，所述在所述可选的集合链路中，选择一条最终链路进行路由转发包括：
    在所述可选的集合链路中，通过轮询RR算法，在具有相同转发能力的集合链路中，随机选择一条链路进行路由转发。
  4. 根据权利要求3所述的方法，其中，所述RR算法包括加权轮询WRR算法、差额轮询DRR算法、紧急轮询URR算法。
  5. 根据权利要求2所述的方法，其中，所述匹配的位置处于任意相邻的两级交换芯片之间。
  6. 一种交换网组播路由的实现装置，所述装置包括：
    获取模块，配置为获取满足转发条件的一类目标转发芯片集合；
    链路建立模块，配置为根据获取的所述满足转发条件的一类目标转发芯片集合，建立可选的集合链路；
    路由选择模块，配置为在所述可选的集合链路中，选择一条最终链路进行路由转发。
  7. 根据权利要求6所述的装置，其中，所述获取模块，配置为在相邻 的两级交换芯片之间，采用匹配基准找到满足条件式
    Figure PCTCN2016082149-appb-100002
    的一类目标转发芯片集合。
  8. 根据权利要求6所述的装置，其中，所述路由选择模块，配置为在所述可选的集合链路中，通过轮询RR算法，在具有相同转发能力的集合链路中，随机选择一条链路进行路由转发。
  9. 根据权利要求8所述的装置，其中，所述RR算法包括加权轮询WRR算法、差额轮询DRR算法、紧急轮询URR算法。
  10. 根据权利要求7所述的装置，其中，所述匹配的位置处于任意相邻的两级交换芯片之间。
  11. 一种计算机存储介质，其中存储有计算机可执行指令，该计算机可执行指令配置执行上述权利要求1至5任一项所述的交换网组播路由的实现方法。
PCT/CN2016/082149 2015-08-18 2016-05-13 一种交换网组播路由的实现方法和装置、计算机存储介质 WO2017028564A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201510507072.XA CN106470161A (zh) 2015-08-18 2015-08-18 一种交换网组播路由的实现方法和装置
CN201510507072.X 2015-08-18

Publications (1)

Publication Number Publication Date
WO2017028564A1 true WO2017028564A1 (zh) 2017-02-23

Family

ID=58050698

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2016/082149 WO2017028564A1 (zh) 2015-08-18 2016-05-13 一种交换网组播路由的实现方法和装置、计算机存储介质

Country Status (2)

Country Link
CN (1) CN106470161A (zh)
WO (1) WO2017028564A1 (zh)

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102238072A (zh) * 2010-05-06 2011-11-09 中兴通讯股份有限公司 一种动态选择路由的方法及clos交换网系统
CN104243323A (zh) * 2013-06-06 2014-12-24 中兴通讯股份有限公司 交换网多播路由方法及系统
US20140376546A1 (en) * 2013-06-20 2014-12-25 The Boeing Company Switch Routing Algorithm

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102238072A (zh) * 2010-05-06 2011-11-09 中兴通讯股份有限公司 一种动态选择路由的方法及clos交换网系统
CN104243323A (zh) * 2013-06-06 2014-12-24 中兴通讯股份有限公司 交换网多播路由方法及系统
US20140376546A1 (en) * 2013-06-20 2014-12-25 The Boeing Company Switch Routing Algorithm

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
CAO, GUODONG.: "Research on Clos Network Multicast Switching Technology", CHINA MASTER THESES FULL-TEXT DATABASE INFORMATION TECHNOLOGY SERIES, 15 March 2013 (2013-03-15), pages 28, 29, ISSN: 1674-0246 *
SHI, ZENGZENG.: "Research on Routing Algorithms in Clos Matrix", CHINA MASTER'S THESES FULL-TEXT DATABASE INFORMATION TECHNOLOGY SERIES, 15 July 2009 (2009-07-15), pages 23, 24, ISSN: 1674-0246 *

Also Published As

Publication number Publication date
CN106470161A (zh) 2017-03-01

Similar Documents

Publication Publication Date Title
US10237206B1 (en) Equal cost multiple path group failover for multicast
WO2018232490A1 (en) TRANSACTION TRAFFIC FORMAT BASED ON A MULTINIVE QUEUE FOR BLOCK CHAINS
US20140071987A1 (en) Systems and methods providing reverse path forwarding compliance for a multihoming virtual routing bridge
EP3560148B1 (en) Database functions-defined network switch
KR20150016309A (ko) 팻-트리 라우팅에 기반하여 별개의 인피니밴드 서브넷들 사이에서 트래픽을 라우팅하기 위한 시스템 및 방법
CN101834789B (zh) 面向包-电路交换片上路由器的回退转向路由算法及所用路由器
EP3559833B1 (en) Best-efforts database functions
CN105052092B (zh) 在包含虚拟网络设备的网络环境中优化路径的方法、网络设备和计算机可读介质
US20130286894A1 (en) Method and Device for Processing Data Cell
CN110891019A (zh) 一种基于负载均衡的数据中心流量调度方法
CN105357124B (zh) 一种MapReduce带宽优化方法
US20160043902A1 (en) Topology discovery in a stacked switches system
US10237205B2 (en) Switch routing algorithms
EP3442172B1 (en) Network topology system and building methods for topologies and routing tables thereof
CN104022950B (zh) 一种可共享和自配置缓存的路由器结构
CN105072032B (zh) 一种确定片上网络路由路径的方法及系统
US9185026B2 (en) Tagging and synchronization for fairness in NOC interconnects
CN108234310A (zh) 多层次互连网络、自适应路由方法及路由设备
CA3223804A1 (en) Deadlock-free multipath routing for direct interconnect networks
CN105704180B (zh) 数据中心网络的配置方法及其系统
WO2015192483A1 (zh) 一种路由节点、路由交换方法、系统及计算机存储介质
WO2015176243A1 (zh) 一种改进型环状拓扑结构及其应用方法
CN106209518B (zh) 一种基于“包-电路”交换技术的动态转向路由算法
WO2017084228A1 (zh) 软件定义网络中流量项目的管理方法
WO2017174019A1 (zh) 一种路由信息处理方法、分组交换设备及存储介质

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: 16836429

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: 16836429

Country of ref document: EP

Kind code of ref document: A1