WO2009015574A1 - Trame d'échange, routeur de mise en grappes - Google Patents

Trame d'échange, routeur de mise en grappes Download PDF

Info

Publication number
WO2009015574A1
WO2009015574A1 PCT/CN2008/071613 CN2008071613W WO2009015574A1 WO 2009015574 A1 WO2009015574 A1 WO 2009015574A1 CN 2008071613 W CN2008071613 W CN 2008071613W WO 2009015574 A1 WO2009015574 A1 WO 2009015574A1
Authority
WO
WIPO (PCT)
Prior art keywords
switching
frame
unit
switch
cascading
Prior art date
Application number
PCT/CN2008/071613
Other languages
English (en)
French (fr)
Inventor
Dajun Zang
Lu Yang
Wenhua Du
Dongcheng Yang
Lingqiang Fan
Gang Gai
Da Zhou
Zhengjie Pu
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 EP08783693A priority Critical patent/EP2175587A4/en
Publication of WO2009015574A1 publication Critical patent/WO2009015574A1/zh
Priority to US12/691,341 priority patent/US8743733B2/en

Links

Classifications

    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L12/00Data switching networks
    • H04L12/54Store-and-forward switching systems 
    • H04L12/56Packet switching systems
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L45/00Routing or path finding of packets in data switching networks
    • H04L45/58Association of routers
    • H04L45/583Stackable routers
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L49/00Packet switching elements
    • H04L49/15Interconnection of switching modules
    • H04L49/1515Non-blocking multistage, e.g. Clos
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L49/00Packet switching elements
    • H04L49/45Arrangements for providing or supporting expansion

Definitions

  • the present invention relates to a switch frame, and more particularly to a switch frame having a multi-plane structure, which is suitable for expanding a cluster router without replacing a switch frame.
  • the present invention also relates to a cluster router, and more particularly to a switch frame having the foregoing switch frame.
  • a cluster router can be expanded without replacing the switch chassis.
  • IP Internet Protocol
  • 3rd Generation third-generation mobile communication
  • video conferencing video on demand
  • P2P point-to-point
  • Clustering is one of the most effective techniques for solving scalability problems.
  • the introduction of clustering technology in the field of routers aims to connect two or more common core routers in a certain way, enabling these core routers to work together between devices.
  • parallel processing to achieve a smooth expansion of system capacity, and externally only as a logical router.
  • parallel switching technology Parallel Packet Switch, PPS for short
  • PPS Parallel Packet Switch
  • a scheme of setting a central switching frame is generally adopted, and a central switching frame is used to implement data exchange between the forwarding frames.
  • various manufacturers generally adopt a multi-plane switching structure.
  • the first solution of the prior art includes four forwarding blocks T (Routing Nodes), and each of the forwarding frames T has five independent switching network boards Ts, which are numbered 0-4.
  • the TX-Matrix platform also includes five independent switching network boards Ts, which are also numbered 0-4. Between these switching stencils Ts-corresponding connections, in fact, constitutes five independent switching planes, each of which is distributed in one switching frame TX and four forwarding frames T.
  • the external network is connected to the ingress packet forward engine (Ingres s PFE) through a physical interface card (Physical Interface Card), and is processed by the ingress packet to the engine, and then distributed to each source in units of sources. Exchange plane.
  • each secondary switching network board Ts in the switching frame TX has only four interfaces, so Four forwarding frames T are cascaded. If you want to cascade it with more forwarding frames T, you must replace the entire switching frame TX.
  • the switching frame fails, the data services of the four forwarding frames connected thereto are completely interrupted. Therefore, the system reliability of the cluster router is Lower.
  • FIG. 2 it is a schematic diagram of a 9-line card frame cascading structure in a cluster router formed by multi-chassis cascading in a second scheme of the prior art.
  • the interior of each switching plane is a 3-level switching network.
  • the first and third level switching network boards Ts l, 3 of each switching plane are set in the forwarding frame; and the second level switching network board Ts 2 is set in the central switching frame TX.
  • the forwarding frame T is composed of a switching network board Ts 2, an electrical backplane Tb, and a plurality of line cards T1.
  • a fiber crossover unit TXa is disposed between the switching network boards Ts l, 3 of each switching plane and the switching network board Ts 2.
  • the fiber cross-connect unit TXa is responsible for rearranging the cascaded fibers between the switch frame TX and the forwarding frame T and then accessing the second-stage switching network board Ts 2 of the switch frame TX.
  • the fiber reassembly can be realized by the fiber crossover unit TXa, and the connection with more forwarding frames T can be realized without replacing the second-stage switching network board Ts 2, thereby realizing the upgrade of the cluster router.
  • each fiber crossover unit TXa corresponds to one level 2 switching network board Ts l, 3.
  • Each fiber crossover unit TXa is provided with nine fiber optic cable interfaces, and each fiber optic cable interface is connected to one forwarding frame T.
  • a fiber-optic crossover unit TXa is bundled with a switching network board Ts 2 to provide nine optical cable interfaces, which are responsible for the exchange between the nine forwarding frames T.
  • the entire 9-frame cascade system requires 8 such binding units.
  • each fiber crossover unit TXa with 18 fiber optic cable interfaces is respectively bound with two switching network boards Ts 2 to form a binding unit.
  • Each binding unit provides 18 optical cable interfaces, which are responsible for data exchange between 18 forwarding frames T.
  • the structure in which the foregoing binding unit is cascaded with 18 forwarding frames T is the same as that shown in FIG. 2.
  • the entire 18-frame cascaded system requires eight such binding units, and the switching capacity of each bonded unit is doubled compared to the nine-frame cascaded system.
  • the fiber crossover unit with 18 fiber optic cable interfaces needs to be replaced with a fiber crossover unit with 36 fiber optic cable interfaces, and each fiber crossover unit and 4 blocks respectively.
  • the switching between the switching frame TX and the 36 forwarding frames T can be implemented by the switching network board Ts 2 binding.
  • a first aspect of the present invention provides a type of switching frame by using a cascading unit of any one of the cascading units of the switching box and a switching port of any switching unit, which can support no replacement of any switching frame. Under the premise of components, the expansion of cluster router capacity is realized.
  • a second aspect of the present invention provides a cluster router of the foregoing structure by using another embodiment, the cluster router including the switch frame having the foregoing structure, and without replacing any switch frame components.
  • the expansion of the capacity of the cluster router can also reduce the expansion cost.
  • Some embodiments of the first aspect of the present invention provide a type of switching frame, which is composed of more than one cascading unit and one or more switching unit connections; wherein the cascading unit is provided with a cascading connection for forwarding frames
  • the switching unit is provided with a switching port for connecting to the cascade interface; any one of the cascaded interfaces of any of the cascaded units is connected to one of the switching units of any of the switching units.
  • each cascade interface of the cascading unit is connected to all switching units; each switching unit is also connected to all cascading interfaces.
  • the switching unit can implement any exchange between multiple cascaded interfaces. It provides technical conditions for the cluster router to expand without replacing parts, which can reduce the expansion cost.
  • Some embodiments of the second aspect of the present invention provide a cluster router, where the cluster router includes a switching frame and a forwarding frame, and the switching frame and the forwarding frame are connected by an optical cable; the switching frame includes more than one cascading unit and one or more switching units.
  • the connection unit is configured; the cascade unit is provided with a cascade interface for connecting the forwarding frame; and any cable interface of any forwarding frame is uniquely connected to one cascade interface of any cascade unit; the exchange unit is provided for connecting the cascade The switch port of the interface; and any cascade interface of any cascade unit is connected to one switch port of any switch unit.
  • the cluster router of the second aspect of the present invention is constructed, and the optical switch connection relationship between all the switching frames and all the forwarding frames can be changed by adding a new switching frame.
  • the switch frame in the foregoing two types of embodiments of the present invention can conveniently form a cluster router together with other switch frames of the same type. Therefore, even if one of the switch frames fails, the other switch frames that are faulty can still work.
  • the connected forwarding frame can also complete its task of forwarding data packets, and the data service will not be interrupted. Therefore, its reliability is significantly higher than the two technical solutions in the aforementioned background art.
  • FIG. 1 is a schematic structural diagram of a cluster router in the prior art
  • FIG. 2 is a schematic structural diagram of another cluster router in the prior art
  • FIG. 3 is a schematic structural diagram of a capacity expansion switch frame in the prior art shown in FIG. 2;
  • Figure 4 is a schematic structural view of an embodiment of the first aspect of the present invention.
  • FIG. 5 is a schematic diagram of connection of an embodiment of a cluster router according to a second aspect of the present invention.
  • FIG. 6 is a schematic diagram of connection of an embodiment of a cluster router expanded on the basis of FIG. 5;
  • FIG. 7 is another connection diagram of an embodiment of a cluster router expanded on the basis of FIG. 5. detailed description
  • the switching frame of the first aspect of the present invention includes p switching units TXC, where the value of p is a positive integer; each switching unit TXC has mxn switching ports; and further includes m cascading units INTC, The values of m and n are also positive integers.
  • each switching unit TXC has a switching network chip, and all switching network chips in each switching unit TXC have m x n switching ports, and therefore, each switching unit has m x n ports TXP.
  • m x n level connection ports INT may be provided. Any cascading interface I NT is connected to a switching port TXP of any switching network chip.
  • the specific connection method is: Any cascading unit Any one of the cascading interfaces of the INTC is connected to a switching port TXP of any switching unit TXC.
  • the number of cascading units of the switching frame is two, and the number of cascading interfaces in each cascading unit is two; the number of switching units is two, and each switching unit
  • the number of switch ports TXP is 2 X 2; the switch ports of each switch unit are divided into groups with 2 switch ports, and the number of switch ports TXP in each group of switch ports is 2, each cascade interface
  • the connection to each switch port can be as follows:
  • the first way the first cascade interface of the first cascade unit and the first of all exchange units a first one of the switch ports of the first switch port, and a second switch port of the first group of switch ports of the first switch unit are respectively connected to the second switch port of the first group of switch ports of the switch unit;
  • the first cascading interface is respectively connected with the first one of the second group of switching ports of all switching units, the second cascading interface of the second cascading unit and the second group of switching ports of all switching units respectively The second exchange port in the connection.
  • the way of this connection is a way of orthogonal connection, that is: each cascading interface of all cascading units is connected to the switching port corresponding to the switching unit orthogonally to the cascading unit.
  • the second mode the first cascading interface of the first cascading unit is respectively connected to one switching port of all switching units, and the connection manner is not the foregoing orthogonal mode, but the following conditions are met:
  • the first cascading interface is only connected to one of the switching ports of any one of the switching units; for example: the first cascading interface of the first cascading unit and the second one of the first group of switching ports of the first switching unit The port is connected and connected to the first one of the second switching ports of the second switching unit.
  • the second cascading interface of the first cascading unit is respectively connected to one switching port of all switching units, and the connection manner thereof is not the foregoing orthogonal mode, but the following conditions are satisfied: the second level of the first cascading unit
  • the joint interface is respectively connected with a switch port of any one of the switch units that is not connected to the first cascade interface of the first cascade unit; for example: the second cascade interface of the first cascade unit and the first switch unit first
  • the first switch port in the group switch port is connected and connected to the second switch port in the second switch port of the second switch unit.
  • the foregoing number of cascading units and switching units are positive integers, respectively.
  • the number of cascading units and switching units can be two. This makes it more advantageous to connect the connection between the cascading unit and the switching unit in the switch frame in an orthogonal manner.
  • the switching network chip set on the switching unit in the switching box determines the number of switching ports on the switching unit, that is, the number of switching ports of all switching network chips that the switching unit has.
  • the quantity is the number of switching ports of the entire switching unit.
  • the switching frame in this embodiment is composed of a number P of switching boards (i.e., switching units) and a number of forwarding boards (i.e., cascading units), wherein each switching unit has ⁇ ⁇ switching ports, each The forwarding board has a cascade interface; the values of ⁇ and ⁇ are respectively a power of 2 (ie 1, 2, 4, 8, 16, 327), and the value of ⁇ is a positive integer.
  • the forwarding board is responsible for connecting to the forwarding frame, and the switching board is responsible for data exchange between the various cascade interfaces.
  • each switch board and each forwarding board in the switch fabric are connected to each other.
  • Each switching board is provided with a switching network chip.
  • the number of switching network chips on each switching board is one. However, in some large-capacity switching boards, the use is not excluded.
  • a plurality of switching network chips form an array of switching networks.
  • each switching network chip has a plurality of switching ports.
  • each row is a connection correspondence between a corresponding cascading interface of the corresponding number forwarding board and each corresponding switching port of each switching board.
  • each optical cable interface of the N forwarding frames is connected to a cascade interface corresponding to the N forwarding boards, and N forwarding frames are implemented.
  • the connection of the P switch boards in the exchange box of the foregoing embodiment constitutes a complete cluster router.
  • FIG. 5 is a schematic diagram showing the connection relationship between the cascading interfaces and the forwarding frames in the switching frame TX according to an embodiment of the second aspect of the present invention.
  • the ⁇ X ⁇ cascade ports in the switch box ⁇ and the corresponding numbered fiber optic cable interfaces in the forwarding box are connected to each other.
  • each switch board in the switch frame is connected to each of the forwarding boards in an orthogonal manner or in a non-orthogonal manner, as long as the connection manner of each embodiment of the switch box is met, as shown in FIG.
  • the connection relationship constitutes the corresponding cluster router.
  • the switch box is interconnected with 2 forwarding frames.
  • the specific connection manner is as follows:
  • the two switching frames are respectively referred to as a first switching frame and a second switching frame; according to the switching plane number, the first switching frame and the second switching frame both have an odd plane and an even plane. .
  • the forwarding frame connected to the even plane of the first switching frame is switched to the odd plane of the second switching frame; for the newly added forwarding frames, respectively connected to the first switching frame and The even plane of the second switching frame; this completes the expansion of the cluster router, and expands its capacity by a factor of two.
  • the foregoing embodiment is only one of a plurality of expansion connection modes. If the forwarding frame connected to the even number plane of the first switching frame is switched to the even number of the second switching frame for the original forwarding frame For the newly added N forwarding frames, it is also possible to connect half of them to the even plane of the first switching frame and the other half to the odd plane of the second switching frame.
  • the original forwarding frame can be connected to the second switching frame to implement the foregoing expansion.
  • FIG. 7 is a connection diagram of an example of a cascading of a forwarding frame, the corresponding manner is the same as that of the foregoing embodiment, and details are not described herein again.

Landscapes

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

Description

交换框、 集群路由器 技术领域
本发明涉及一种交换框, 特别是一种具有多平面结构的交换框, 适用于 不更换交换框而使集群路由器扩容; 本发明还涉及一种集群路由器, 特别是 一种具有前述交换框的集群路由器, 可以不更换交换框实现扩容。 背景技术
随着宽带网络的飞速发展,尤其是基于宽带网络的各类新业务层出不穷, 未来对网络的带宽要求也会呈高速发展的趋势。 网际协议 ( Internet Protocol, 简称: IP) 电话、 第三代移动通信( 3rd Generation, 简称: 3G ) 业务、 视频会议、 视频点播以及大量新兴的基于点到点 (Point to Point, 简称: P2P) 的业务应用, 都在迅速消耗着承载网络的剩余带宽。
集群是一种最有效的解决扩展性问题的技术, 集群技术在路由器领域的 引入, 目的是将两台或两台以上普通核心路由器通过某种方式连接, 使得这 些核心路由器能够进行设备间协同工作和并行处理, 实现系统容量的平滑扩 展, 并且对外只表现为一台逻辑路由器。 通过釆用并行交换技术(Parallel Packet Switch, 简称: PPS ), 将多个独立的交换网级联, 共同组成一个多级 多平面的交换矩阵系统, 从而突破单机箱在交换容量、 功耗、 散热等方面的 限制, 实现更大容量的路由交换系统。
目前在路由器等通信设备通过机框间级联方式实现容量扩展时, 普遍釆 用设置中央交换框的方案,用中央交换框来实现各个转发框之间的数据交换。 在中央交换框的实现方法上, 各个厂家普遍釆用多平面交换结构。
参见图 1, 现有技术的第一种方案包括 4个转发框 T (Routing Node), 每个转发框 T 中设有 5 个独立的交换网板 Ts, 其编号为 0-4。 交换框 TX (TX-Matrix platform )也包括 5个独立的交换网板 Ts , 其编号也是 0-4。 这些交换网板 Ts之间——对应连接, 实际上构成了 5个独立的交换平面, 每 个交换平面都分布在 1个交换框 TX和 4个转发框 T里。 需要说明的是: 图 1 中只显示了 2个转发框 T , 另外的两个转发框 T与交换框 TX的连接关系与显 示的两个转发框 T和交换框 TX的连接关系是一样的, 故在图 1中省略。 外部 网络通过物理接口卡 (Phys i ca l Interface Card , 简称: PIC )接入到入口 分组前向引擎( Ingres s PFE ), 经过该入口分组前向引擎处理后, 以信源为 单位分配到各个交换平面。
本发明的发明人在发明过程中发现, 前述方案在实际的应用中, 可扩展 性不好, 具体表现在该交换框 TX中每个二级交换网板 Ts只有 4个接口, 故 只能与 4个转发框 T进行级联。 若想使其与更多的转发框 T进行级联, 就必 须换掉整个交换框 TX。 另外, 由于前述的集群路由器中只有 1个交换框 ΤΧ, 一旦该交换框 ΤΧ发生故障,则与其连接的 4个转发框 Τ的数据业务就会全部 中断, 因此, 这种集群路由器的系统可靠性较低。
参见图 2 , 其为现有技术的第二种方案通过多框级联构成的集群路由器 中 9线卡框级联结构示意图。 其中, 包括 8个交换平面, 这 8个平面之间均 衡地分担负载。 每个交换平面的内部是一个 3级交换网络。 在物理实现上, 把每个交换平面的第 1、 3级交换网板 Ts l,3设置在转发框 Τ中; 把第 2级交 换网板 Ts 2设置在中央交换框 TX上。 第 1级到第 2级之间, 第 2级到第 3级 之间, 分别釆用框间级联光缆进行连接。 转发框 T 由交换网板 Ts 2、 电背板 Tb以及若干线卡 T1构成。
与第一种方案不同的是:该结构中,在每一个交换平面的交换网板 Ts l, 3 和交换网板 Ts 2之间都设置了一个光纤交叉单元 TXa。 光纤交叉单元 TXa 负 责把交换框 TX和转发框 T之间的级联光纤重新整理后再接入到交换框 TX的 第 2级交换网板 Ts 2。 通过光纤交叉单元 TXa可以实现光纤重组, 在不更换 第 2级交换网板 Ts 2的情况下就可实现与更多的转发框 T连接, 从而实现集 群路由器的升级。 在上述第二种方案中, 每个光纤交叉单元 TXa要对应 1个 2级交换网板 Ts l,3。 每个光纤交叉单元 TXa设有 9个光缆接口, 每个光缆接口连接 1个转 发框 T。 1个光纤交叉单元 TXa与 1个交换网板 Ts 2绑定在一起, 可以提供 9 个光缆接口, 负责 9个转发框 T之间的交换。 整个 9框级联系统则需要 8套这 样的绑定单元。
参见图 3 , 当前述的集群路由器容量升级到 18个转发框 T进行级联时, 需 要将交换框 TX中原具有 9个光缆接口的光纤交叉单元 TXa更换为至少具有 18个 光缆接口的光纤交叉单元 TXa , 每个具有 18个光缆接口的光纤交叉单元 TXa分 别与 2块交换网板 Ts 2绑定, 构成一个绑定单元。 每个绑定单元提供 18个光缆 接口, 负责分别 18个转发框 T之间进行的数据交换。前述绑定单元与 18个转发 框 T级联的结构, 与图 2所示的相同。 整个 18框级联的系统需要 8套这样的绑 定单元, 与前述 9框级联的系统相比, 每个绑定单元的交换容量都增加了 1倍。
依前所述, 当需要将集群路由器的容量继续扩充时, 需将原具有 18个光 缆接口的光纤交叉单元再换成具有 36个光缆接口的光纤交叉单元,每个光纤 交叉单元分别与 4块交换网板 Ts 2绑定, 就可以实现交换框 TX与 36个转发 框 T之间的数据交换。
前述第二种技术方案虽然能够实现对了集群路由器的容量升级, 但是, 由于升级时需要更换交换框的光纤交叉单元, 而光纤交叉单元的成本较高, 因此会导致用户投资的浪费。 发明内容
本发明的第一个方面是通过一些实施例提供一类交换框, 该种交换框的 级联单元任一级联接口和任一交换单元的一个交换端口连接, 可以支持在不 更换任何交换框部件的前提下, 实现集群路由器容量的扩充。
本发明的第二个方面是通过另一些实施例提供一类集群路由器, 该集群 路由器包括具有前述结构的交换框,是现在不更换任何交换框部件的前提下 , 对集群路由器容量的扩充, 同时可以降低扩容成本。
本发明的第一个方面的一些实施例提供了一类交换框, 该交换框, 由一 个以上级联单元和一个以上交换单元连接构成; 其中, 级联单元设有用于连 接转发框的级联接口; 交换单元设有用于连接所述级联接口的交换端口; 任 一级联单元的任一级联接口和任一交换单元的一个交换端口连接。
本发明前述实施例的交换框, 其级联单元的每个级联接口连接到所有的 交换单元; 每个交换单元也连接到所有的级联接口。 通过这种连接结构, 交 换单元可以实现多个级联接口之间的任意交换。 为集群路由器能够在不更换 部件的前提下的扩容提供了技术条件, 能够降低扩容成本。
本发明第二个方面的一些实施例提供了一种集群路由器, 该集群路由器 包括交换框和转发框, 交换框和转发框通过光缆连接; 该交换框由一个以上 级联单元和一个以上交换单元连接构成; 级联单元设有用于连接转发框的级 联接口; 且任一转发框的任一光缆接口唯一地和任一级联单元的一个级联接 口连接; 交换单元设有用于连接级联接口的交换端口; 且任一级联单元的任 一级联接口和任一交换单元的一个交换端口连接。
在前述第一方面实施例的基础上,构造本发明第二个方面的集群路由器, 并且, 可以通过增加新的交换框, 并通过改变所有的交换框和所有的转发框 之间的光缆连接关系, 可以实现 N个, 2N个, 4N个直到 Μ χ Ν个转发框之间的任 意的交换, 其中, Ν和 Μ是 2的冪数, 即: N=2 M=2j , i和 j为 0或者正整数。 因此, 可以在集群路由器扩容时, 只增加新的交换框和转发框部件, 而不需 要替换任何已有的部件, 降低了扩容成本。
另外, 本发明前述两类实施例中的交换框可以方便地与其他同类型交换 框共同构成集群路由器, 因此, 即使其中的一个交换框发生故障, 其他为发 生故障的交换框依然能够工作, 与其连接的转发框也就能够完成其转发数据 包的任务, 数据业务不会中断。 因此, 其可靠性相比于前述背景技术中的两 种技术方案明显较高。 附图说明
图 1为现有技术中一种集群路由器的结构示意图;
图 2为现有技术中另一种集群路由器的结构示意图;
图 3为图 2所示现有技术中扩容交换框的结构示意图;
图 4为本发明第一个方面的实施例的结构示意图;
图 5为本发明第二个方面集群路由器的实施例的连接示意图;
图 6为在图 5基础上扩容的集群路由器实施例的连接示意图;
图 7为在图 5基础上扩容的集群路由器实施例的另一连接示意图。 具体实施方式
图 4为本发明第一个方面的实施例的结构示意图。 参见图 4 , 本发明第 一方面实施例的交换框包括 p个交换单元 TXC, p的取值是正整数; 每个交换 单元 TXC具有 m x n个交换端口; 此外, 还包括 m个级联单元 INTC, m、 n的 取值也是正整数。 在每个级联单元上有设有 n个级联接口 INT, 每个级联接 口对应设有 p个通道。 实际上, 每个交换单元 TXC中都设有交换网芯片, 每 个交换单元 TXC中所有的交换网芯片具有 m x n个交换端口, 因此, 每个交换 单元具有 m x n个端口 TXP。 本实施例的一个交换框中可设有 m x n个级联接 口 INT。 任何一个级联接口 I NT与任何一个交换网芯片的一个交换端口 TXP 连接。 具体的连接方式是: 任一级联单元 INTC的任一级联接口 I NT和任一交 换单元 TXC的一个交换端口 TXP连接。
例如, 假设本发明一实施例中, 交换框的级联单元数量为 2个, 每个级 联单元中的级联接口的数量均为 2; 交换单元的数量为 2个, 每个交换单元 的交换端口 TXP的数量均为 2 X 2个;将每个交换单元的交换端口都分为具有 2个交换端口的组, 每组交换端口中的交换端口 TXP数量都是 2个, 各个级 联接口和各个交换端口之间的连接可以有如下的方式:
第一种方式: 第一级联单元的第一个级联接口分别和所有交换单元的第 一组交换端口中的第一个交换端口连接, 第一级联单元的第二个级联接口分 别和所有交换单元的第一组交换端口中的第二个交换端口连接; 第二级联单 元的第一个级联接口分别和所有交换单元的第二组交换端口中的第一个交换 端口连接, 第二级联单元的第二个级联接口分别和所有交换单元的第二组交 换端口中的第二个交换端口连接。 这种连接的方式是一种正交连接的方式, 即: 所有级联单元的各个级联接口都连接到交换单元与该级联单元正交对应 的那一个交换端口上。
第二种方式: 第一级联单元的第一个级联接口分别连接到所有交换单元 的一个交换端口, 其连接方式不是前述的正交方式, 但满足如下的条件: 第 一级联单元的第一个级联接口分别只和任一交换单元中的一个交换端口连 接; 例如: 第一级联单元的第一个级联接口和第一交换单元第一组交换端口 中的第二个交换端口连接, 并和第二交换单元第二组交换端口中的第一个交 换端口连接。 第一级联单元的第二个级联接口分别连接到所有交换单元的一 个交换端口, 其连接方式也不是前述的正交方式, 但满足如下的条件: 第一 级联单元的第二个级联接口分别和任一交换单元的一个未与第一级联单元的 第一级联接口连接的交换端口连接; 例如: 第一级联单元的第二个级联接口 和第一交换单元第一组交换端口中的第一个交换端口连接, 并和第二交换单 元第二组交换端口中的第二个交换端口连接。
上述第二种方式随着交换框中交换单元和级联单元数量的增加, 其变化 形式也会有多种, 但无论如何变化都要满足: 任一级联单元的任一级联接口 和任一交换单元的一个交换端口连接。
一般情况下, 前述的级联单元和交换单元的数目分别是正整数。 在具体 设置时, 可以使级联单元和交换单元的数量都是 2的养数。 这样可以更加有 利于将交换框中级联单元和交换单元之间的连接方式釆用正交方式连接。
需要说明的是: 交换框中的交换单元上设置的交换网芯片决定了该交换 单元上的交换端口数量, 即交换单元具有的所有交换网芯片的交换端口的数 量即是整个交换单元的交换端口的数量。
为了更加详细地说明前述较为概括性的实施例, 以正交连接方式为例, 再次详细介绍本发明的又一个具体的实施例。 在该实施例中的交换框由数量 为 P的交换板(即: 交换单元)和数量为 M的转发板(即级联单元)构成, 其 中, 每一个交换单元具有 Μχ Ν交换端口, 每个转发板具有 Ν个级联接口; Μ 和 Ν的取值分别是 2的冪数(即 1, 2, 4, 8, 16, 32...... ), Ρ的取值是正整 数。 转发板负责连接转发框, 交换板负责各个级联接口之间的数据交换。 为 了达到交换的目的, 交换框中的每个交换板和每个转发板之间都相互连接。 每个交换板上都设有交换网芯片, 在本实施例中, 每个交换板上的交换网芯 片的数量都是 1个, 但是, 在某些较大容量的交换板中, 不排除使用多个交 换网芯片构成交换网阵列。
依前所述, 在本实施例中, 每个交换网芯片有 Μχ Ν个交换端口。 交换框 中有 Ρ个交换板, 每个交换板上有 Μχ Ν个交换端口。 所以一个交换框中共有 Μχ Ν个级联接口。 这 Μ X Ν个级联接口与 Ρ个交换网芯片中每一个交换网芯 片的 Μ X Ν个交换端口之间——对应连接。当需要 Ν个转发板在交换框互连时, 转发板和交换板的连接对应关系如表 1所示:
表 1 转发板 级联接口 交换端口
1 11 11
1 12 12
1
1M 1M
Ν 1 N1
Ν 2 Ν2
Ν
Μ ΝΜ 表 1 中, 转发板对应的数字、 符号表示各个转发板的编号, 转发板级联 接口对应的数字、 符号表示各个转发板级联接口的编号; 交换端口对应的数 字、 符号表示每个交换板的交换端口的编号。 该表 1 中, 各行分别是相应编 号转发板的相应级联接口与各个交换板相应各个交换端口的连接对应关系。 按照表 1连接完成后, 假设整个交换框共有 P个交换板, 则该交换框提供了 P个交换平面。 交换框中的每个交换板都实现了与 N个转发板之间的连接。
参见图 5 , 本发明第二方面一些实施例是在前述交换框的基础上, 再将 N 个转发框的各个光缆接口和 N个转发板对应的级联接口连接, 就实现了 N个 转发框和前述实施例交换框中 P个交换板的连接,构成一完整的集群路由器。 图 5所示是本发明第二方面一个实施例的交换框 TX中 Ν χ Μ个级联接口和 Ν 个转发框对应连接关系的示意图。 其中的交换框 ΤΧ中的 Ν X Μ个级联端口和 Ν个转发框中对应编号的光缆接口相互连接。
依前所述, 无论交换框内部各个交换板与各个转发板釆用正交方式连接 或釆用非正交方式连接, 只要符合前述交换框各个实施例的连接方式即可釆 用图 5所示的连接关系构成相应的集群路由器。
参见图 6 , 在前述交换框的基础上, 本发明第二个方面的另外一些实施 例如下: 当需要对前述集群路由器进行扩容时, 例如: 容量扩大一倍, 可以 利用两个前述实施例的交换框与 2Ν个转发框互连。 具体的连接方式如下: 为了描述上方便, 将两个交换框分别称作第一交换框和第二交换框; 按 照交换平面编号, 第一交换框和第二交换框都具有奇数平面和偶数平面。
对于原 Ν个转发框, 将其中连接到第一交换框偶数平面的转发框换接到 第二交换框的奇数平面; 对于新增的 Ν个转发框, 将其分别连接到第一交换 框和第二交换框的偶数平面; 这样就完成了对集群路由器的扩容, 使其容量 扩大到原来的 2倍。
前面的实施例仅仅是众多扩容连接方式的一种,如果对于原 Ν个转发框, 将其中连接到第一交换框偶数数平面的转发框换接到第二交换框的偶数平 面; 对于新增的 N个转发框, 将其一半连接到第一交换框的偶数平面, 另一 半连接到第二交换框的奇数平面也是可以的。
此外, 只要保证新增的转发框能够连接到第一交换框, 原来的转发框能 够连接到第二交换框也可以实现前述的扩容。
实际上, 按照前述的方式, 最多可以实现 Ν χ Μ个转发框的级联。 当 Ν χ Μ个转发框级联时, 级联的系统中应当包括 Μ个交换框和 Ν χ Μ个转发框。 每 个交换框 Μ X Ν个交换端口分别为唯一地与 Ν X Μ个转发框上相应的光缆接口 ——连接。 参见图 7 , 其为 Ν χ Μ个转发框的级联的一个实例的连接示意图, 其对应方式与前述实施例是一样的, 在此不再赘述。
最后应说明的是:以上各个实施例仅用以说明本发明的技术方案而非对 本发明进行限制, 尽管参照上述各个实施例对本发明的主要技术方案进行了 详细说明, 本领域的普通技术人员应当理解: 其依然可以在本发明前述各个 实施例的技术方案基础上进行修改或者等同替换; 而这些修改或者等同替换 并不脱离本发明各个实施例所揭示的技术方案的精神和范围。

Claims

权 利 要求
1、 一种交换框, 其特征在于: 该交换框由一个以上级联单元和一个以上 交换单元连接构成;
所述级联单元设有用于连接转发框的级联接口;
所述交换单元设有用于连接所述级联接口的交换端口;
任一级联单元的任一级联接口和任一交换单元的一个交换端口连接。
2、 根据权利要求 1所述的交换框, 其特征在于: 所述任一级联单元的任 一级联接口釆用正交方式和任一交换单元的一个交换端口连接。
3、 根据权利要求 1或 2所述的交换框, 其特征在于: 所述的级联单元为 M个, 且 M=2j, j为 0或者正整数。
4、 根据权利要求 3所述的交换框, 其特征在于: 每个所述的级联单元具 有 N个级联接口, 且 N=2 i为 0或者正整数。
5、 根据权利要求 3所述的交换框, 其特征在于: 每个所述交换单元的交 换端口为 M XN个, 且 N=2 i为 0或者正整数。
6、 一种集群路由器, 包括交换框和转发框, 交换框和转发框通过光缆连 接; 其特征在于:
所述交换框由一个以上级联单元和一个以上交换单元连接构成; 所述级联单元设有用于连接转发框的级联接口; 且任一转发框的任一光 缆接口唯一地和任一级联单元的一个级联接口连接;
所述交换单元设有用于连接所述级联接口的交换端口; 且任一级联单元 的任一级联接口和任一交换单元的一个交换端口连接。
7、 根据权利要求 6所述的集群路由器, 其特征在于: 所述级联单元的任 一级联接口釆用正交方式和任一交换单元的一个交换端口连接。
8、 根据权利要求 6或 7所述的集群路由器, 其特征在于: 所述的级联单 元为 M个, 且 M=2j, j为 0或者正整数。
9、 根据权利要求 8所述的集群路由器, 其特征在于: 每个所述的级联单 元具有 N个级联接口, 且 N=2 i为 G或者正整数。
10、 根据权利要求 8所述的集群路由器, 其特征在于: 每个所述的交换 单元为 M X N个, 且 N=2', i为 0或者正整数。
PCT/CN2008/071613 2007-07-27 2008-07-11 Trame d'échange, routeur de mise en grappes WO2009015574A1 (fr)

Priority Applications (2)

Application Number Priority Date Filing Date Title
EP08783693A EP2175587A4 (en) 2007-07-27 2008-07-11 EXCHANGE FRAME, ROUTING ROUTER
US12/691,341 US8743733B2 (en) 2007-07-27 2010-01-21 Switching frame and router cluster

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN2007101289606A CN101355430B (zh) 2007-07-27 2007-07-27 交换框、集群路由器
CN200710128960.6 2007-07-27

Related Child Applications (1)

Application Number Title Priority Date Filing Date
US12/691,341 Continuation US8743733B2 (en) 2007-07-27 2010-01-21 Switching frame and router cluster

Publications (1)

Publication Number Publication Date
WO2009015574A1 true WO2009015574A1 (fr) 2009-02-05

Family

ID=40303890

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2008/071613 WO2009015574A1 (fr) 2007-07-27 2008-07-11 Trame d'échange, routeur de mise en grappes

Country Status (4)

Country Link
US (1) US8743733B2 (zh)
EP (1) EP2175587A4 (zh)
CN (1) CN101355430B (zh)
WO (1) WO2009015574A1 (zh)

Families Citing this family (17)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101631081B (zh) 2009-08-12 2011-06-08 华为技术有限公司 一种多级交换网
CN102217252B (zh) * 2009-08-14 2013-10-09 华为技术有限公司 集群设备的通信系统
WO2011057447A1 (zh) * 2009-11-13 2011-05-19 华为技术有限公司 路由器及集群路由器
JP5521258B2 (ja) * 2010-03-29 2014-06-11 ▲ホア▼▲ウェイ▼技術有限公司 クラスタルータおよびクラスタルーティング方法
CN101895398B (zh) 2010-07-15 2012-07-25 华为技术有限公司 数据通信方法和装置
CN101917337B (zh) * 2010-08-09 2015-06-03 中兴通讯股份有限公司 路由器集群中板间互联的装置及方法
WO2011113394A2 (zh) * 2011-04-27 2011-09-22 华为技术有限公司 一种路由器、虚拟集群路由器系统及建立方法
CN103534995B (zh) 2011-08-19 2016-01-13 华为技术有限公司 一种路由器集群的板间通信方法、路由器和路由器集群
CN103780516B (zh) 2012-10-19 2017-06-09 华为技术有限公司 一种立体背板
US9398354B2 (en) * 2013-02-20 2016-07-19 Xtera Communications, Inc. Integrated assembly for switching optical signals
CN104641593B (zh) * 2013-03-01 2018-02-13 华为技术有限公司 网板和通信设备
CN105245454B (zh) * 2014-07-10 2018-10-19 华为技术有限公司 交换系统的流量转发方法和装置
CN104902352A (zh) * 2015-05-14 2015-09-09 烽火通信科技股份有限公司 一种多业务框、交换框的大容量集群设备
CN107239427B (zh) 2017-06-09 2020-12-01 华为技术有限公司 连接装置、网板连接系统和通信设备
CN107547255B (zh) * 2017-06-30 2020-12-01 新华三技术有限公司 集群系统的升级方法、主用主控板及集群系统
CN109286564B (zh) * 2017-07-20 2022-06-07 迈普通信技术股份有限公司 一种报文转发方法及装置
CN107682274A (zh) * 2017-10-27 2018-02-09 上海力诺通信科技有限公司 一种基于以太网的正交架构平台

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
DE4228694A1 (de) 1992-08-28 1994-03-03 Siemens Ag Verfahren zum Durchschalten von Multiplexsignalen
US7110394B1 (en) 2001-06-25 2006-09-19 Sanera Systems, Inc. Packet switching apparatus including cascade ports and method for switching packets
CN1983841A (zh) * 2006-05-11 2007-06-20 华为技术有限公司 一种级连系统及实现备份的方法

Family Cites Families (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US3842214A (en) * 1972-10-31 1974-10-15 Stromberg Carlson Corp Path finding system
LU87976A1 (de) * 1991-03-14 1992-04-15 Siemens Ag Dreistufige,zumindest gedoppelte atm-koppelanordnung
US5321813A (en) * 1991-05-01 1994-06-14 Teradata Corporation Reconfigurable, fault tolerant, multistage interconnect network and protocol
US5481073A (en) * 1994-06-09 1996-01-02 Quintech, Inc. Modular broadband bidirectional programmable switch system with stacked modular switch arrangement
US5878177A (en) * 1997-03-31 1999-03-02 At&T Corp. Layered switch architectures for high-capacity optical transport networks
US6591285B1 (en) * 2000-06-16 2003-07-08 Shuo-Yen Robert Li Running-sum adder networks determined by recursive construction of multi-stage networks
AU2002255763A1 (en) * 2001-03-16 2002-10-03 Photuris, Inc. Modular all-optical cross-connect
US6590417B1 (en) * 2001-04-03 2003-07-08 Cypress Semiconductor Corporation Cascadable bus based crossbar switch in a programmable logic device
US7006438B2 (en) * 2001-05-31 2006-02-28 Turin Networks Distributed control of data flow in a network switch
US7899334B2 (en) * 2003-01-31 2011-03-01 Ciena Corporation Signal distribution module for a directionless reconfigurable optical add/drop multiplexer
US7224184B1 (en) * 2004-11-05 2007-05-29 Xilinx, Inc. High bandwidth reconfigurable on-chip network for reconfigurable systems

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
DE4228694A1 (de) 1992-08-28 1994-03-03 Siemens Ag Verfahren zum Durchschalten von Multiplexsignalen
US7110394B1 (en) 2001-06-25 2006-09-19 Sanera Systems, Inc. Packet switching apparatus including cascade ports and method for switching packets
CN1983841A (zh) * 2006-05-11 2007-06-20 华为技术有限公司 一种级连系统及实现备份的方法

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
See also references of EP2175587A4 *

Also Published As

Publication number Publication date
EP2175587A4 (en) 2010-07-28
CN101355430B (zh) 2012-02-29
US20100118867A1 (en) 2010-05-13
EP2175587A1 (en) 2010-04-14
CN101355430A (zh) 2009-01-28
US8743733B2 (en) 2014-06-03

Similar Documents

Publication Publication Date Title
WO2009015574A1 (fr) Trame d'échange, routeur de mise en grappes
US20230052529A1 (en) Network interconnect as a switch
CN106789679B (zh) 一种线卡框、多框集群路由器、选路及报文处理方法
JP4843087B2 (ja) スイッチング帯域幅を向上させるスイッチングシステムおよび方法
JP3163069B2 (ja) スイッチ構造のアップグレード方法
TWI233018B (en) Multi-port high-speed serial fabric interconnect chip in a meshed configuration
JP6605747B2 (ja) ラインカードシャーシ、マルチシャーシクラスタルータおよびパケット処理
WO2007129397A1 (ja) ビデオ信号スイッチングシステム
CN102326358B (zh) 一种集群系统扩容方法、装置及集群系统
WO2012119408A1 (zh) 一种多框级联装置
EP2521326B1 (en) Run-time scalable switching fabric
WO2017143940A1 (zh) 一种交换方法及交换系统
CN100389569C (zh) 一种数据通信装置
CN1859175A (zh) 内部可靠互连的通信设备
KR20050020583A (ko) 다차원의 절단된 그물 스위칭 네트워크
CN103269285A (zh) 网络通信集群装置和用于实现网络通信集群的方法
CN101232510A (zh) 一种多网口设备组播的实现方法及系统
CN115988363A (zh) 一种光线路终端olt中数据通信的方法和olt
WO2021159936A1 (zh) 连接结构及设备安装方法
JP2016503975A (ja) マルチキャストトラフィックブリッジング
US11265265B2 (en) Data communications system and method
CN100352227C (zh) 多插框互联系统
JP2008187236A (ja) 通信システムおよび通信方法
CN100571186C (zh) 实现动态evc多粒度电路交换网络的方法
CN116132364A (zh) 报文传输方法、装置及系统

Legal Events

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

Ref document number: 08783693

Country of ref document: EP

Kind code of ref document: A1

WWE Wipo information: entry into national phase

Ref document number: 2008783693

Country of ref document: EP

NENP Non-entry into the national phase

Ref country code: DE