WO2014169691A1 - 一种信元数据处理方法、装置和交换网元 - Google Patents

一种信元数据处理方法、装置和交换网元 Download PDF

Info

Publication number
WO2014169691A1
WO2014169691A1 PCT/CN2013/090991 CN2013090991W WO2014169691A1 WO 2014169691 A1 WO2014169691 A1 WO 2014169691A1 CN 2013090991 W CN2013090991 W CN 2013090991W WO 2014169691 A1 WO2014169691 A1 WO 2014169691A1
Authority
WO
WIPO (PCT)
Prior art keywords
cell data
multicast
load balancing
port number
network element
Prior art date
Application number
PCT/CN2013/090991
Other languages
English (en)
French (fr)
Inventor
赵姣
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 WO2014169691A1 publication Critical patent/WO2014169691A1/zh

Links

Classifications

    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L47/00Traffic control in data switching networks
    • H04L47/10Flow control; Congestion control
    • H04L47/12Avoiding congestion; Recovering from congestion
    • H04L47/125Avoiding congestion; Recovering from congestion by balancing the load, e.g. traffic engineering

Definitions

  • the present invention relates to the field of data communications, and in particular, to a cell data processing method, apparatus, and switching network element. Background technique
  • dedicated switching network elements are usually connected in a single-stage or multi-level interconnection to form a switching network to achieve large data exchange capacity requirements.
  • the data is sent from a source access chip and transmitted through a multi-stage switching network composed of multiple switched network element chips to the destination access chip.
  • a multi-stage switching network composed of multiple switched network element chips to the destination access chip.
  • unicast cell data For data transmitted in a switched network, it can be generally divided into unicast cell data and multicast cell data.
  • the unicast cell data is not copied, and is transmitted through the switching network, and the source access chip is one-to-one to the destination access chip; the multicast cell data is copied in the switching network element chip, and transmitted through the switching network.
  • the multicast should be replicated as much as possible on the switching network element chip closest to the destination access chip.
  • the switching network element chip constituting the switching network system needs to perform path selection and the like in the process of transmitting the unicast and multicast cell data to the destination end.
  • the application number is CN201110202414.9, and the patent document of the cell processing method and device for the switching network element chip proposes a cell processing method, which specifically includes:
  • a unicast routing table for the unicast cell in the switching network element chip the content is the unicast identification ID and the output port number correspondence; constructing a multicast routing table for the multicast cell, the content is the multicast group identification ID Corresponding to the output port number; constructing a port aggregation trunk table for the multicast cell, the content is the correspondence between the output port number and the trunk group (the port number interconnected with the downstream chip of the same chip is divided into one For the members of the trunk group, for the sake of simplicity, each trunk group contains fixed members and the number is equal. The content of each trunk group is the port number of the same downstream chip connected to the switching network.
  • the unicast routing table, the multicast routing table, and the trunk table are shown in FIG. 2, FIG. 3, and FIG. 4, respectively.
  • the switching network element chip receives unicast cell data and multicast cell data from an upstream chip (access chip or switched network element chip). For the unicast cell data, the switching network element chip obtains the output port number bit map according to the unicast ID query unicast routing table; according to whether there is back pressure on the output link, the output port number bit map corresponding bit is erased to 0; According to the output port number bitmap of the wipe 0, all available output ports are polled and selected to obtain the final output port. For the multicast cell data, the switching network element chip queries the multicast routing table according to the multicast group identifier carried by it to obtain the output port number bitmap; according to whether there is back pressure on the output link, the output port number bit map corresponding bit is erased.
  • the switching network element chip obtains the output port number to which the multicast cell is to be copied, and ensures that the multicast cell is even if there is a multi-port connection with the downstream chip. Only copy one copy to the chip.
  • the switching network element chip obtains the output port number of the unicast cell through the unicast routing table and the polling circuit, and obtains the copy output port of the multicast cell through the multicast routing table, the trunk table, and the polling circuit. number.
  • the problem with this method is that the path selection between the output ports is realized only by simple polling scheduling. It is proved by experiments that the load balancing effect is not good, and the switching network congestion caused by uneven traffic distribution is easily caused. This is especially true when the switched network delivers variable length cells. Summary of the invention
  • an object of the present invention to provide a method, a device and a switching network element for processing a cell data, which are capable of load balancing all optional links of cell data according to the cumulative amount of data of a plurality of links included in the switching network element.
  • an embodiment of the present invention provides a method for processing a cell data, which is used to exchange network elements, where the switching network element has a unicast routing table corresponding to unicast cell data and multicast cell data. Corresponding multicast routing table;
  • the method includes:
  • the method for processing a cell data wherein, when the cell data is unicast cell data, the destination identifier of the cell data is a unicast identifier;
  • the unicast identifier in the unicast routing table corresponds to an output port number.
  • the method for processing a cell data wherein, when the cell data is multicast cell data, the destination identifier of the cell data is a multicast group identifier;
  • the multicast group identifier in the multicast routing table corresponds to the device number of the downstream device.
  • the switching network element further has a multicast load balancing table corresponding to the multicast cell data, the device number of the downstream device in the multicast load balancing table, and the switching network element.
  • the link number of the link corresponds to
  • Determining, according to the destination identifier of the cell data, a routing table corresponding to the output port number corresponding to the cell data querying a multicast routing table according to the destination identifier of the multicast cell data
  • the multicast load balancing table obtains an output port number bitmap corresponding to the multicast cell data.
  • the load balancing is performed on the output port number bitmap according to the accumulated data amount of the plurality of links included in the switching network element, and the load balancing result is obtained by:
  • the port number corresponding to the link with the smallest accumulated data amount is selected as the output destination port number.
  • the load balancing of the output port number bitmap is performed according to the accumulated data amount of the plurality of links included in the switching network element, and the load balancing result further includes:
  • the current cache depth value of the link whose accumulated data amount reaches a preset threshold is obtained;
  • the current buffer depth value of the link whose accumulated amount of data reaches a preset threshold resets the accumulated amount of the data amount of the link.
  • an embodiment of the present invention further provides a cell data processing apparatus, which is located in a switching network element, where the switching network element has a unicast routing table corresponding to unicast cell data and multicast cell data. Corresponding multicast routing table;
  • the device includes:
  • the receiving module is configured to: receive, by the switching network element, the cell data that carries the destination identifier from the upstream device;
  • Obtaining a module configured to: query a corresponding routing table according to the destination identifier of the cell data, and obtain an output port number bitmap corresponding to the cell data;
  • the load balancing processing module is configured to: perform load balancing on the output port number bitmap according to the accumulated data volume of the multiple links included in the switching network element, to obtain a load balancing result; and a destination port number determining module, setting And: determining, according to the load balancing result, an output destination port number, and outputting the cell data according to the destination port number.
  • the destination identifier of the cell data is a unicast identifier
  • the unicast identifier in the unicast routing table corresponds to an output port number.
  • the above-described cell data processing device wherein, when the cell data is multicast cell data,
  • the destination identifier of the metadata is a multicast group identifier
  • the multicast group identifier in the multicast routing table corresponds to the device number of the downstream device.
  • the switching network element further has a multicast load balancing table corresponding to the multicast cell data, the device number of the downstream device in the multicast load balancing table, and the switching network element
  • the link number of the link corresponds to
  • the obtaining module is configured to: query a multicast routing table and a multicast load balancing table according to the destination identifier of the multicast cell data, and obtain an output port number bitmap corresponding to the multicast cell data.
  • the load balancing processing module includes: a first calculating module, configured to: calculate an accumulated amount of data amount of each link among the plurality of links included in the switching network element;
  • the load balancing processing sub-module is configured to: obtain the load balancing result according to the accumulated data amount.
  • the destination port number determining module includes: a selecting module, configured to: select a port number corresponding to the link with the smallest accumulated data amount as the output destination port number.
  • the load balancing processing module further includes: a determining module, configured to: determine whether a cumulative value of the data volume of each of the plurality of links reaches a preset threshold Obtain the judgment result;
  • a second calculation module configured to: when the judgment result indicates that the accumulated data value of any one of the multiple links reaches a preset threshold, acquiring a link of the data volume cumulative value reaching a preset threshold Current cache depth value;
  • the reset module is configured to: reset the accumulated data amount of the link by using a current cache depth value of the link whose accumulated data amount reaches a preset threshold.
  • the embodiment of the present invention further provides a switching network element, where the switching network element has a unicast routing table corresponding to unicast cell data, a multicast routing table corresponding to the multicast cell data, And a multicast load balancing table corresponding to the multicast cell data;
  • the switching network element further includes the cell data processing apparatus of any of the above.
  • the embodiment of the present invention it is possible to accumulate data according to the data volume of multiple links included in the switching network element. All the optional links of the cell data are load-balanced. Further, the problem of uneven traffic distribution caused by poor load balancing effect can be solved.
  • the flexibility of the connection relationship between the chips and the good load balancing effect can be realized while completing the data processing of the switching network.
  • 1 is an application scenario of a data processing device of a large-scale switching network in the related art
  • FIG. 5 is a schematic diagram of a method for processing load balancing by using a polling schedule for multicast cell data in the related art
  • FIG. 6 is a schematic flowchart of a method for processing a cell data according to an embodiment of the present invention
  • FIG. 7 is a multicast routing table according to an embodiment of the present invention.
  • FIG. 8 is a multicast load balancing table according to an embodiment of the present invention.
  • FIG. 9 is a schematic structural diagram of a multicast load balancing scheduling circuit according to an embodiment of the present invention
  • FIG. 10 is a schematic structural diagram of a 4-link multicast load balancing circuit according to an embodiment of the present invention
  • An embodiment of the present invention provides a method for processing a cell data, which is used to exchange network elements, where the switching network element has a unicast routing table corresponding to unicast cell data and a multicast route corresponding to the multicast cell data.
  • the method is as shown in FIG. 6, and includes:
  • Step 61 The switching network element receives the cell data carrying the destination identifier from the upstream device.
  • Step 62 Query the corresponding routing table according to the destination identifier of the cell data, and obtain an output port number corresponding to the cell data.
  • Bit map Step 63 Perform load balancing on the output port number bitmap according to the accumulated data volume of the multiple links included in the switching network element, to obtain a load balancing result.
  • Step 64 Determine an output destination port number according to the load balancing result, and output the cell data according to the destination port number.
  • the upstream device herein may be an access chip or a switching network element chip, according to the cell data.
  • the destination identifier is used to query a routing table corresponding to the cell data, and obtains a bitmap of the output port number of the cell data.
  • the output port number bit is The figure performs load balancing to obtain a load balancing result, and determines an output destination port number according to the load balancing result, and outputs the cell data according to the destination port number.
  • load balancing is performed on all the optional links of the cell data based on the accumulated data volume of the multiple links included in the switching network element, and further, the uneven traffic distribution caused by the poor load balancing effect is solved. The problem.
  • the cell data may be unicast cell data or multicast cell data.
  • the cell data processing method provided by the embodiment of the present invention is described below for different cell data.
  • the destination identifier of the cell data is a unicast identifier; and the unicast identifier in the unicast routing table corresponds to an output port number.
  • the unicast routing table is a routing table corresponding to the unicast identifier and the output port number, which is the same as the unicast routing table in the prior art, as shown in FIG. 2 .
  • the method for processing the cell data includes: Step 1: Query the unicast routing table according to the unicast identifier carried by the unicast identifier to obtain a unicast output port number bitmap;
  • Step 2 Perform load balancing on the unicast output port number bitmap according to the accumulated data volume of the multiple links included in the switching network element, and obtain a load balancing result.
  • Step 3 Determine, according to the load balancing result, an output destination port number, and output the unicast cell data according to the destination port number.
  • the destination identifier of the cell data is a multicast group identifier
  • the multicast group identifier in the multicast routing table corresponds to the device number of the downstream device.
  • the switching network element further has a multicast load balancing table corresponding to the multicast cell data, and the device number of the downstream device and the chain of the switched network element link in the multicast load balancing table.
  • Determining, according to the destination identifier of the cell data, a routing table corresponding to the output, and obtaining an output port number bitmap corresponding to the cell data specifically, querying a multicast routing table according to the destination identifier of the multicast cell data, and The multicast load balancing table obtains an output port number bitmap corresponding to the multicast cell data.
  • a new type of multicast routing table is constructed.
  • the multicast group identifies a routing table corresponding to the device number of the downstream device, and the downstream device refers to a downstream chip, which may be a switching network element chip or Access to the chip.
  • the multicast load balancing table corresponding to the multicast cell data is further included in the switching network element, as shown in FIG. 8, wherein the device number of the downstream device corresponds to the link number of the switched network element link. .
  • the cell data processing method includes: Step 1: query the multicast routing table according to the multicast group identifier carried by the multicast cell group to obtain a multicast downstream device number bit map;
  • Step 2 query the multicast load balancing table according to the multicast downstream device number bit map obtained in step 1, and obtain a bitmap of the multicast output port number;
  • Step 3 Perform load balancing on the multicast output port number bitmap according to the accumulated data volume of the multiple links included in the switching network element, and obtain a load balancing result.
  • Step 4 Determine, according to the load balancing result, an output destination port number, and output the multicast cell data according to the destination port number.
  • the output port number of the multicast chip is in all the output ports of the switching network element. It can be flexibly selected, so that the connection relationship between the chips and the number of connection ports are no longer fixed, which makes the fiber connection and communication antenna PCB layout and wiring more convenient.
  • step 63 can be specifically:
  • the accumulated data amount of each link is an accumulated value of the data amount of the output port of the link, and the data signal is added whenever the data cell of the output port of the link enters the output end of the switching network element for buffering.
  • step 64 is specifically:
  • the port number corresponding to the link with the smallest accumulated data amount is selected as the output destination port number.
  • the method for processing a cell data provided by the embodiment of the present invention calculates a data amount cumulative value of each link among the multiple links included in the switching network element as a load balancing parameter, and selects a link with the smallest accumulated data amount.
  • the corresponding port number is used as the output destination port number.
  • the step 63 specifically includes:
  • the current cache depth value of the link whose accumulated data amount reaches a preset threshold is obtained;
  • the current buffer depth value of the link whose accumulated amount of data reaches a preset threshold resets the accumulated amount of the data amount of the link.
  • the current cache depth value of the link is used to reset the link.
  • the cumulative amount of data is used to reset the link.
  • a load balancing circuit can be implemented as shown in FIG. 9 to implement load balancing on the output port number bitmap, and a data amount is accumulated for each link of the multiple links included in the network element chip.
  • the counter is used to count the accumulated value of the data volume of the corresponding link.
  • the counter has an upper limit of counting.
  • the accumulated value of the data amount accumulation counter is preferably reset by using the current cache depth value of the corresponding link.
  • a buffer depth counter is set for each link to calculate the current cache depth value of the corresponding link.
  • the buffer depth counter reflects the real-time depth of the link output buffer, that is, the total amount of data selected via the link but not yet output from the output port, whenever the cell data via the link output port enters the switching network.
  • the output end of the meta chip is buffered, and the buffer depth counter increases the amount of data corresponding to the cell data; when the cell data is output from the output port of the link, the buffer depth counter subtracts the amount of data corresponding to the cell data.
  • the load balancing circuit further includes a detection module, configured to detect the data amount accumulation counter of all the output ports in real time, and reset the values of all the output port data amount accumulation counters to when any of the accumulation counters is full.
  • the output port buffers the current value of the depth counter.
  • the load balancing circuit further includes a comparison module, wherein the comparison module is configured to implement a load balancing comparison function of the load balancing, that is, compare the current accumulated counter values of all active links by using multi-stage pipeline mode, and select the smallest value from the selected one.
  • the port number corresponding to the link is the output destination port number of the data cell.
  • the following takes the 4 link as an example to illustrate the multicast cell data load balancing process.
  • a multicast cell data has a total of four links of links 0 to 3, and the accumulated counter values of the four links are compared in two stages, and the link with the smallest accumulator value is selected as the link.
  • the output port number corresponding to the link with the smallest accumulator value is the output destination port number.
  • the buffer depth counter is not used as a basis for comparison.
  • the accumulating counter value is full, the value of the current buffer depth counter is assigned to the accumulating counter, and then the buffer depth counter and the accumulating counter continue to perform respective addition and subtraction operations (the accumulating counter does not do Subtraction operation), does not affect each other.
  • the embodiment of the present invention further provides a cell data processing apparatus, configured to exchange a network element, where the switching network element has a unicast routing table corresponding to unicast cell data and a multicast corresponding to the multicast cell data. Routing table
  • the device is shown in Figure 11, and includes:
  • a receiving module configured to receive, by the switching network element, the cell data that carries the destination identifier from the upstream device, and the acquiring module, configured to query the corresponding routing table according to the destination identifier of the cell data, to obtain An output port number bitmap corresponding to the cell data;
  • a load balancing processing module configured to perform load balancing on the output port number bitmap according to the data volume accumulation situation of the multiple links included in the switching network element, to obtain a load balancing result
  • a destination port number determining module configured to determine an output destination port number according to the load balancing result, and output the cell data according to the destination port number.
  • the destination identifier of the cell data is a unicast identifier
  • the unicast identifier in the unicast routing table corresponds to an output port number.
  • the destination identifier of the cell data is a multicast group identifier
  • the multicast group identifier in the multicast routing table corresponds to the device number of the downstream device.
  • the switching network element further has a multicast load balancing table corresponding to the multicast cell data, the device number of the downstream device in the multicast load balancing table, and the switching network element
  • the link number of the link corresponds to
  • the acquiring module is configured to query a multicast routing table and a multicast load balancing table according to the destination identifier of the multicast cell data, and obtain an output port number bitmap corresponding to the multicast cell data.
  • the load balancing processing module specifically includes: a first calculating module, configured to calculate an accumulated amount of data of each link of the plurality of links included in the switching network element;
  • the load balancing processing submodule is configured to obtain the load balancing result according to the accumulated amount of the data amount.
  • the destination port number determining module specifically includes: a selecting module, configured to select a port number corresponding to the link with the smallest accumulated data amount as the output destination port number.
  • the load balancing processing module further includes: a determining module, configured to determine whether a data amount cumulative value of each link of the plurality of links reaches a preset threshold Obtain the judgment result;
  • a second calculating module configured to: when the judgment result indicates that the accumulated data value of any one of the multiple links reaches a preset threshold, obtain a link that the accumulated data amount reaches a preset threshold Pre-cache depth value;
  • a reset module configured to reset the accumulated data amount of the link by using a current buffer depth value of the link whose accumulated data amount reaches a preset threshold.
  • the embodiment of the present invention further provides a switching network element, where the switching network element has a unicast routing table corresponding to unicast cell data, a multicast routing table corresponding to multicast cell data, and a multicast message.
  • Multicast load balancing table corresponding to metadata
  • the switching network element further includes the cell data processing apparatus of any of the above.
  • the flexibility of the connection relationship between the chips and the good load balancing effect can be realized while completing the data processing of the switching network.

Landscapes

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

Abstract

一种信元数据处理方法、装置和交换网元,其中,所述方法包括:所述交换网元从上游设备接收携带目的标识的信元数据;根据所述信元数据的目的标识查询对应的路由表,获得与所述信元数据对应的输出端口号比特图;根据所述交换网元包括的多条链路的数据量累计情况,对所述输出端口号比特图进行负载均衡,得到负载均衡结果;根据所述负载均衡结果,确定输出目的端口号,并将所述信元数据按所述目的端口号输出。本发明实施例能够根据交换网元包括的多条链路的数据量累计情况对信元数据的所有可选链路进行负载均衡。

Description

一种信元数据处理方法、 装置和交换网元
技术领域
本发明涉及数据通信领域, 尤其涉及一种信元数据处理方法、 装置和交 换网元。 背景技术
在数据通讯领域, 特别是大规模交换网系统的设计中, 通常将专用的交 换网元芯片釆用单级或多级互联的方式联接起来构成交换网络, 以达到大的 数据交换容量要求。
如图 1所示, 数据从某片源接入芯片发出, 经多片交换网元芯片组成的 多级交换网络传输之后,到达目的接入芯片。通常为保证链路连接的健壮性, 需要提供一定的冗余度或增加芯片间的连接带宽, 因此, 在每两片交换网元 芯片及接入芯片之间会有多条链路连接。
对于在交换网中传输的数据来说, 一般可分为单播信元数据和多播信元 数据两种。 其中单播信元数据不进行复制, 经交换网传输, 由源接入芯片一 对一的去往目的接入芯片; 多播信元数据在交换网元芯片中进行复制, 经交 换网传输, 由源接入芯片一对多的去往目的接入芯片。 为了緩解交换网传输 中的拥塞, 在多级级联构建交换网时, 多播应尽量在最靠近目的接入芯片的 交换网元芯片进行复制。
构成交换网系统的交换网元芯片在将单播和多播信元数据传输到目的端 的过程中需要对其进行路径选择等一系列处理。申请号为 CN201110202414.9, 名称为交换网元芯片的信元处理方法及装置的专利文件提出了一种信元处理 办法, 具体包括:
在交换网元芯片中为单播信元构建一张单播路由表, 内容为单播标识 ID 与输出端口号对应关系; 为多播信元构建一张多播路由表, 内容为多播组标 识 ID与输出端口号对应关系; 为多播信元构建一张端口聚合 trunk表, 内容 为输出端口号与 trunk组的对应关系 (与同片下游芯片互联的端口号划为一 个 trunk组的成员, 为简便起见, 每个 trunk组所包含的成员固定且数目均相 等) , 每个 trunk组内的内容为与交换网片连接的同片下游芯片的端口号。 其中, 所用的单播路由表、 多播路由表、 trunk表分别如图 2、 图 3、 图 4所 示。
交换网元芯片从上游芯片 (接入芯片或交换网元芯片)接收单播信元数 据和多播信元数据。 对于单播信元数据, 交换网元芯片根据其携带的单播 ID 查询单播路由表获得输出端口号比特图; 根据输出链路是否存在反压, 将输 出端口号比特图对应位抹 0; 根据抹 0后的输出端口号比特图, 对所有可用 输出端口进行轮询选择, 得到最终的输出端口。 对于多播信元数据, 交换网 元芯片根据其携带的多播组标识查询多播路由表获得输出端口号比特图; 根 据输出链路是否存在反压, 将输出端口号比特图对应位抹 0; 根据抹 0后的 输出端口号比特图, 查询 trunk表获得其所属的 trunk组, 在同 trunk组内多 个可用输出端口间进行轮询选择, 得到最终的输出端口, 如图 5所示, 当输 出端口号比特图从属于多个不同的 trunk组时, 最终获得的输出端口数等于 从属的 trunk组数。 通过多播路由表和 trunk表的查询, 交换网元芯片获得多 播信元要复制去往的输出端口号, 且确保即使在与下游同片芯片存在多端口 连接的情况下, 多播信元往该芯片只复制一份。
釆用上述处理方法, 交换网元芯片通过单播路由表及轮询电路获得单播 信元的输出端口号, 通过多播路由表、 trunk表及轮询电路获得多播信元的复 制输出端口号。 但此种方法存在的问题在于: 仅通过简单的轮询调度来实现 输出端口间的路径选择, 经实验证实其负载均衡效果不佳, 并容易导致由于 流量分配不均而产生的交换网络拥塞, 这在交换网络传递变长信元时尤为突 出。 发明内容
本发明的目的是提供一种信元数据处理方法、 装置和交换网元, 能够根 据交换网元包括的多条链路的数据量累计情况对信元数据的所有可选链路进 行负载均衡。 为了实现上述目的, 本发明实施例提供了一种信元数据处理方法, 用于 交换网元, 所述交换网元具有与单播信元数据对应的单播路由表和与多播信 元数据对应的多播路由表;
所述方法包括:
所述交换网元从上游设备接收携带目的标识的信元数据;
根据所述信元数据的目的标识查询对应的路由表, 获得与所述信元数据 对应的输出端口号比特图;
根据所述交换网元包括的多条链路的数据量累计情况, 对所述输出端口 号比特图进行负载均衡, 得到负载均衡结果;
根据所述负载均衡结果, 确定输出目的端口号, 并将所述信元数据按所 述目的端口号输出。
上述的信元数据处理方法, 其中, 所述信元数据为单播信元数据时, 所 述信元数据的目的标识为单播标识;
所述单播路由表中单播标识与输出端口号对应。
上述的信元数据处理方法, 其中, 所述信元数据为多播信元数据时, 所 述信元数据的目的标识为多播组标识;
所述多播路由表中多播组标识与下游设备的设备号对应。
上述的信元数据处理方法, 其中, 所述交换网元还具有与多播信元数据 对应的多播负载均衡表, 多播负载均衡表中所述下游设备的设备号与所述交 换网元链路的链路号对应;
所述根据所述信元数据的目的标识查询对应的路由表, 获得与所述信元 数据对应的输出端口号比特图为: 根据所述多播信元数据的目的标识查询多 播路由表以及多播负载均衡表, 获得与所述多播信元数据对应的输出端口号 比特图。
上述的信元数据处理方法, 其中, 所述根据所述交换网元包括的多条链 路的数据量累计情况, 对所述输出端口号比特图进行负载均衡, 得到负载均 衡结果包括:
计算所述交换网元包括的多条链路中每条链路的数据量累计值; 根据所述数据量累计值, 得到所述负载均衡结果。 上述的信元数据处理方法, 其中, 所述根据所述负载均衡结果, 确定输 出目的端口号为:
选择所述数据量累计值最小的链路所对应的端口号作为所述输出目的端 口号。
上述的信元数据处理方法, 其中, 所述根据所述交换网元包括的多条链 路的数据量累计情况, 对所述输出端口号比特图进行负载均衡, 得到负载均 衡结果还包括:
判断所述多条链路中的每条链路的数据量累计值是否达到预设阔值, 获 得判断结果;
当判断结果指示所述多条链路中的任一链路的数据量累计值达到预设阔 值时, 获取所述数据量累计值达到预设阔值的链路的当前緩存深度值; 用所述数据量累计值达到预设阔值的链路的当前緩存深度值重置所述链 路的所述数据量累计值。
为了实现上述目的, 本发明实施例还提供了一种信元数据处理装置, 位 于交换网元, 所述交换网元具有与单播信元数据对应的单播路由表和与多播 信元数据对应的多播路由表;
所述装置包括:
接收模块, 设置为: 所述交换网元从上游设备接收携带目的标识的信元 数据;
获取模块, 设置为: 根据所述信元数据的目的标识查询对应的路由表, 获得与所述信元数据对应的输出端口号比特图;
负载均衡处理模块, 设置为: 根据所述交换网元包括的多条链路的数据 量累计情况, 对所述输出端口号比特图进行负载均衡, 得到负载均衡结果; 目的端口号确定模块, 设置为: 根据所述负载均衡结果, 确定输出目的 端口号, 并将所述信元数据按所述目的端口号输出。
上述的信元数据处理装置, 其中, 所述信元数据为单播信元数据时, 所 述信元数据的目的标识为单播标识;
所述单播路由表中单播标识与输出端口号对应。
上述的信元数据处理装置, 其中, 所述信元数据为多播信元数据时, 所 述信元数据的目的标识为多播组标识;
所述多播路由表中多播组标识与下游设备的设备号对应。
上述的信元数据处理装置, 其中, 所述交换网元还具有与多播信元数据 对应的多播负载均衡表, 多播负载均衡表中所述下游设备的设备号与所述交 换网元链路的链路号对应;
所述获取模块设置为: 根据所述多播信元数据的目的标识查询多播路由 表以及多播负载均衡表,获得与所述多播信元数据对应的输出端口号比特图。
上述的信元数据处理装置, 其中, 所述负载均衡处理模块包括: 第一计算模块, 设置为: 计算所述交换网元包括的多条链路中每条链路 的数据量累计值;
负载均衡处理子模块, 设置为: 根据所述数据量累计值, 得到所述负载 均衡结果。
上述的信元数据处理装置, 其中, 所述目的端口号确定模块包括: 选择模块, 设置为: 选择所述数据量累计值最小的链路所对应的端口号 作为所述输出目的端口号。
上述的信元数据处理装置, 其中, 所述负载均衡处理模块还包括: 判断模块, 设置为: 判断所述多条链路中的每条链路的数据量累计值是 否达到预设阔值, 获得判断结果;
第二计算模块, 设置为: 当判断结果指示所述多条链路中的任一链路的 数据量累计值达到预设阔值时, 获取所述数据量累计值达到预设阔值的链路 的当前緩存深度值;
重置模块, 设置为: 用所述数据量累计值达到预设阔值的链路的当前緩 存深度值重置所述链路的所述数据量累计值。
为了实现上述目的, 本发明实施例还提供了一种交换网元, 所述交换网 元具有与单播信元数据对应的单播路由表、 与多播信元数据对应的多播路由 表、 以及与多播信元数据对应的多播负载均衡表;
所述交换网元还包括上述任一项所述的信元数据处理装置。
本发明实施例具有以下有益效果中的至少一项:
本发明实施例, 能够根据交换网元包括的多条链路的数据量累计情况对 信元数据的所有可选链路进行负载均衡, 进一步地, 能够解决负载均衡效果 不佳导致的流量分配不均的问题;
本发明实施例, 能够在完成交换网数据处理的同时, 实现各芯片联接关 系的灵活性及良好的负载均衡效果。
附图概述
图 1为相关技术中大规模交换网数据处理装置应用场景;
图 2为相关技术中的单播路由表;
图 3为相关技术中的多播路由表;
图 4为相关技术中的多播端口聚合 trunk表;
图 5为相关技术中对于多播信元数据釆用轮询调度实现负载均衡的示意 图 6为本发明实施例提供的信元数据处理方法的流程示意图;
图 7为本发明实施例提供的多播路由表;
图 8为本发明实施例提供的多播负载均衡表;
图 9为本发明实施例提供的多播负载均衡调度电路的结构示意图; 图 10为本发明实施例提供的 4链路的多播负载均衡电路的结构示意图; 图 11为本发明实施例提供的信元数据处理装置的结构示意图。 本发明的较佳实施方式
下面将结合附图及具体实施例进行详细描述。
本发明实施例提供了一种信元数据处理方法, 用于交换网元, 所述交换 网元具有与单播信元数据对应的单播路由表和与多播信元数据对应的多播路 由表;
所述方法如图 6所示, 包括:
步骤 61 , 所述交换网元从上游设备接收携带目的标识的信元数据; 步骤 62, 根据所述信元数据的目的标识查询对应的路由表, 获得与所述 信元数据对应的输出端口号比特图; 步骤 63 , 根据所述交换网元包括的多条链路的数据量累计情况, 对所述 输出端口号比特图进行负载均衡, 得到负载均衡结果;
步骤 64, 根据所述负载均衡结果, 确定输出目的端口号, 并将所述信元 数据按所述目的端口号输出。
本发明实施例提供的信元数据处理方法, 其中, 交换网元从上游设备接 收到携带目的标识的信元数据时, 这里的上游设备可以为接入芯片或交换网 元芯片, 根据信元数据的目的标识查询与信元数据对应的路由表, 获得该信 元数据的输出端口号比特图, 此时, 根据交换网元包括的多条链路的数据量 累计情况, 对所述输出端口号比特图进行负载均衡, 得到负载均衡结果, 并 根据负载均衡结果确定输出目的端口号, 将所述信元数据按所述目的端口号 输出。
本发明实施例, 基于交换网元包括的多条链路的数据量累计情况对信元 数据的所有可选链路进行负载均衡, 进一步地, 解决了负载均衡效果不佳导 致的流量分配不均的问题。
信元数据可以为单播信元数据或多播信元数据, 下面分别针对不同的信 元数据说明本发明实施例提供的信元数据处理方法。
<单播信元数据 >
所述信元数据为单播信元数据时,所述信元数据的目的标识为单播标识; 所述单播路由表中单播标识与输出端口号对应。
在本发明实施例中,单播路由表为单播标识与输出端口号对应的路由表, 与现有技术中的单播路由表相同, 如图 2所示。
对于进入交换网元芯片的单播信元数据, 信元数据处理方法包括: 步骤 1 , 根据其携带的单播标识查询单播路由表获得单播输出端口号比 特图;
步骤 2, 根据所述交换网元包括的多条链路的数据量累计情况, 对单播 输出端口号比特图进行负载均衡, 获得负载均衡的结果;
步骤 3 , 根据所述负载均衡结果, 确定输出目的端口号, 并将所述单播 信元数据按所述目的端口号输出。
<多播信元数据 > 所述信元数据为多播信元数据时, 所述信元数据的目的标识为多播组标 识;
所述多播路由表中多播组标识与下游设备的设备号对应。
此时, 优选地, 所述交换网元还具有与多播信元数据对应的多播负载均 衡表, 多播负载均衡表中所述下游设备的设备号与所述交换网元链路的链路 号对应;
所述根据所述信元数据的目的标识查询对应的路由表, 获得与所述信元 数据对应的输出端口号比特图具体为根据所述多播信元数据的目的标识查询 多播路由表以及多播负载均衡表, 获得与所述多播信元数据对应的输出端口 号比特图。
在本发明实施例中, 构建了新型的多播路由表如图 7所示, 为多播组标 识与下游设备的设备号对应的路由表, 下游设备指下游芯片, 可以为交换网 元芯片或接入芯片。 优选地, 在交换网元中还具有与多播信元数据对应的多 播负载均衡表如图 8所示, 其中所述下游设备的设备号与所述交换网元链路 的链路号对应。
对于进入交换网元芯片的多播信元数据, 信元数据处理方法包括: 步骤 1 , 根据其携带的多播组标识查询多播路由表获得多播下游设备号 比特图;
步骤 2, 根据通过步骤 1获得的多播下游设备号比特图查询多播负载均 衡表, 获得多播输出端口号比特图;
步骤 3 , 根据所述交换网元包括的多条链路的数据量累计情况, 对多播 输出端口号比特图进行负载均衡, 获得负载均衡的结果;
步骤 4, 根据所述负载均衡结果, 确定输出目的端口号, 并将所述多播 信元数据按所述目的端口号输出。
由上述过程可以看出, 在实现多播信元数据查表功能时, 通过构建和使 用新型多播路由表及多播负载均衡表, 多播芯片的输出端口号在交换网元所 有输出端口中可以灵活选择, 从而使芯片间连接关系及连接端口数量不再固 定, 使光纤联接和通信天线 PCB布局布线更为方便。
上述的信元数据处理方法中, 无论信元数据为单播信元数据或多播信元 数据, 步骤 63可以具体为:
计算所述交换网元包括的多条链路中每条链路的数据量累计值; 根据所述数据量累计值, 得到所述负载均衡结果。
每条链路的数据量累计值是经由该链路输出端口的数据量的积累值, 每 当经由该链路输出端口的数据信元进入交换网元的输出端进行緩存时, 增加 该数据信元对应的数据量; 当数据信元经由该链路输出时, 不做任何统计。
相应地, 步骤 64具体为:
选择所述数据量累计值最小的链路所对应的端口号作为所述输出目的端 口号。
本发明实施例提供的信元数据处理方法, 通过计算交换网元包括的多条 链路中每条链路的数据量累计值作为负载均衡的参数, 并选择数据量累计值 最小的链路所对应的端口号作为所述输出目的端口号, 在实现单播及多播路 径选择功能时, 通过釆用一套基于输出端口输出数据量累积的算法实现, 得 到较佳的负载均衡效果, 可以更好的避免由于流量分配不均而产生的交换网 络拥塞, 这一效果在信元长度为变长时尤为明显。
上述的信元数据处理方法, 所述步骤 63具体还包括:
判断所述多条链路中的每条链路的数据量累计值是否达到预设阔值, 获 得判断结果;
当判断结果指示所述多条链路中的任一链路的数据量累计值达到预设阔 值时, 获取所述数据量累计值达到预设阔值的链路的当前緩存深度值; 用所述数据量累计值达到预设阔值的链路的当前緩存深度值重置所述链 路的所述数据量累计值。
本发明实施例优选地, 当网元芯片包括的多条链路中的任一链路的数据 量累计值达到预设阔值时, 用所述链路的当前緩存深度值来重置所述链路的 数据量累计值。
在实际应用中, 可以通过一个负载均衡电路如图 9所示, 实现对所述输 出端口号比特图的负载均衡, 针对网元芯片包括的多条链路的每条链路设置 一个数据量累加计数器, 用于统计对应链路的数据量累计值, 一般的计数器 都存在计数上限, 当任一数据量累加计数器计满, 即达到某一预设阔值时, 需要对该数据量累加计数器进行重置, 在本发明实施例中, 优选地, 利用对 应链路的当前緩存深度值来重置所述数据量累加计数器的累计值。
在负载均衡电路中, 为每条链路设置緩存深度计数器来计算对应链路的 当前緩存深度值。 緩存深度计数器反映了链路输出端緩存的实时深度, 即选 定经由该链路输出但是还没有从输出端口输出的数据的总量, 每当经由该链 路输出端口的信元数据进入交换网元芯片的输出端緩存, 緩存深度计数器增 加该信元数据对应的数据量; 当信元数据从链路的输出端口输出时, 緩存深 度计数器减去该信元数据对应的数据量。
该负载均衡电路还包括检测模块, 该检测模块用于实时对所有输出端口 的数据量累加计数器进行检测, 当任一累加计数器计满时, 则将所有输出端 口数据量累加计数器的值重置为该输出端口緩存深度计数器的当前值。
该负载均衡电路还包括比较模块, 该比较模块用于实现负载均衡的链路 负载比较功能, 即釆用多级流水的方式对所有有效链路的当前累加计数器值 进行比较, 从中选出值最小的链路对应的端口号作为数据信元的输出目的端 口号。
下面以 4链路为例, 说明多播信元数据负载均衡过程。
如图 10所示, 一个多播信元数据有链路 0~3共 4条链路可经由输出, 这四条链路的累加计数器值经过两级比较, 选出累加器值最小的链路作为负 载均衡的结果, 这个累加器值最小的链路对应的输出端口号就是输出目的端 口号。
其中, 緩存深度计数器不作为比较的依据, 当累加计数器值计满时, 将 当前緩存深度计数器的值赋给累加计数器, 然后緩存深度计数器和累加计数 器继续执行各自的加减操作 (累加计数器不做减操作) , 互不影响。
本发明实施例还提供了一种信元数据处理装置, 用于交换网元, 所述交 换网元具有与单播信元数据对应的单播路由表和与多播信元数据对应的多播 路由表;
所述装置如图 11所示, 包括:
接收模块,用于所述交换网元从上游设备接收携带目的标识的信元数据; 获取模块, 用于根据所述信元数据的目的标识查询对应的路由表, 获得 与所述信元数据对应的输出端口号比特图;
负载均衡处理模块, 用于根据所述交换网元包括的多条链路的数据量累 计情况, 对所述输出端口号比特图进行负载均衡, 得到负载均衡结果;
目的端口号确定模块, 用于根据所述负载均衡结果, 确定输出目的端口 号, 并将所述信元数据按所述目的端口号输出。
上述的信元数据处理装置, 其中, 所述信元数据为单播信元数据时, 所 述信元数据的目的标识为单播标识;
所述单播路由表中单播标识与输出端口号对应。
上述的信元数据处理装置, 其中, 所述信元数据为多播信元数据时, 所 述信元数据的目的标识为多播组标识;
所述多播路由表中多播组标识与下游设备的设备号对应。
上述的信元数据处理装置, 其中, 所述交换网元还具有与多播信元数据 对应的多播负载均衡表, 多播负载均衡表中所述下游设备的设备号与所述交 换网元链路的链路号对应;
所述获取模块具体为根据所述多播信元数据的目的标识查询多播路由表 以及多播负载均衡表, 获得与所述多播信元数据对应的输出端口号比特图。
上述的信元数据处理装置, 其中, 所述负载均衡处理模块具体包括: 第一计算模块, 用于计算所述交换网元包括的多条链路中每条链路的数 据量累计值;
负载均衡处理子模块, 用于根据所述数据量累计值, 得到所述负载均衡 结果。
上述的信元数据处理装置, 其中, 所述目的端口号确定模块具体包括: 选择模块, 用于选择所述数据量累计值最小的链路所对应的端口号作为 所述输出目的端口号。
上述的信元数据处理装置, 其中, 所述负载均衡处理模块具体还包括: 判断模块 , 用于判断所述多条链路中的每条链路的数据量累计值是否达 到预设阔值, 获得判断结果;
第二计算模块, 用于当判断结果指示所述多条链路中的任一链路的数据 量累计值达到预设阈值时, 获取所述数据量累计值达到预设阈值的链路的当 前緩存深度值;
重置模块, 用于用所述数据量累计值达到预设阔值的链路的当前緩存深 度值重置所述链路的所述数据量累计值。
本发明实施例还提供了一种交换网元, 所述交换网元具有与单播信元数 据对应的单播路由表、 与多播信元数据对应的多播路由表、 以及与多播信元 数据对应的多播负载均衡表;
所述交换网元还包括上述任一项所述的信元数据处理装置。
以上所述仅是本发明的优选实施方式, 应当指出, 对于本技术领域的普 通技术人员来说, 在不脱离本发明原理的前提下, 还可以做出若干改进和润 饰, 这些改进和润饰也应视为本发明的保护范围。
工业实用性
本发明实施例, 能够在完成交换网数据处理的同时, 实现各芯片联接关 系的灵活性及良好的负载均衡效果。

Claims

权 利 要 求 书
1.一种信元数据处理方法, 用于交换网元, 其中, 所述交换网元具有与 单播信元数据对应的单播路由表和与多播信元数据对应的多播路由表;
所述方法包括:
所述交换网元从上游设备接收携带目的标识的信元数据;
根据所述信元数据的目的标识查询对应的路由表, 获得与所述信元数据 对应的输出端口号比特图;
根据所述交换网元包括的多条链路的数据量累计情况, 对所述输出端口 号比特图进行负载均衡, 得到负载均衡结果;
根据所述负载均衡结果, 确定输出目的端口号, 并将所述信元数据按所 述目的端口号输出。
2. 如权利要求 1所述的信元数据处理方法, 其中, 所述信元数据为单播 信元数据时, 所述信元数据的目的标识为单播标识;
所述单播路由表中单播标识与输出端口号对应。
3. 如权利要求 1所述的信元数据处理方法, 其中, 所述信元数据为多播 信元数据时, 所述信元数据的目的标识为多播组标识;
所述多播路由表中多播组标识与下游设备的设备号对应。
4.如权利要求 3所述的信元数据处理方法, 其中, 所述交换网元还具有 与多播信元数据对应的多播负载均衡表, 所述多播负载均衡表中所述下游设 备的设备号与所述交换网元链路的链路号对应;
所述根据所述信元数据的目的标识查询对应的路由表, 获得与所述信元 数据对应的输出端口号比特图为: 根据所述多播信元数据的目的标识查询多 播路由表以及多播负载均衡表, 获得与所述多播信元数据对应的输出端口号 比特图。
5.如权利要求 1-4任一项所述的信元数据处理方法, 其中, 所述根据所 述交换网元包括的多条链路的数据量累计情况, 对所述输出端口号比特图进 行负载均衡, 得到负载均衡结果包括:
计算所述交换网元包括的多条链路中每条链路的数据量累计值; 根据所述数据量累计值, 得到所述负载均衡结果。
6. 根据权利要求 5所述的信元数据处理方法, 其中,根据所述负载均衡 结果, 确定输出目的端口号为:
选择所述数据量累计值最小的链路所对应的端口号作为所述输出目的端 口号。
7.如权利要求 6所述的信元数据处理方法, 其中, 所述根据所述交换网 元包括的多条链路的数据量累计情况, 对所述输出端口号比特图进行负载均 衡, 得到负载均衡结果还包括:
判断所述多条链路中的每条链路的数据量累计值是否达到预设阔值, 获 得判断结果;
当判断结果指示所述多条链路中的任一链路的数据量累计值达到预设阔 值时, 获取所述数据量累计值达到阔值的链路的当前緩存深度值;
用所述数据量累计值达到预设阔值的链路的当前緩存深度值重置所述链 路的所述数据量累计值。
8.—种信元数据处理装置, 位于交换网元, 其中, 所述交换网元具有与 单播信元数据对应的单播路由表和与多播信元数据对应的多播路由表; 所述装置包括:
接收模块, 设置为: 所述交换网元从上游设备接收携带目的标识的信元 数据;
获取模块, 设置为: 根据所述信元数据的目的标识查询对应的路由表, 获得与所述信元数据对应的输出端口号比特图;
负载均衡处理模块, 设置为: 根据所述交换网元包括的多条链路的数据 量累计情况, 对所述输出端口号比特图进行负载均衡, 得到负载均衡结果; 目的端口号确定模块, 设置为: 根据所述负载均衡结果, 确定输出目的 端口号, 并将所述信元数据按所述目的端口号输出。
9. 如权利要求 8所述的信元数据处理装置, 其中, 所述信元数据为单播 信元数据时, 所述信元数据的目的标识为单播标识;
所述单播路由表中单播标识与输出端口号对应。
10. 如权利要求 8所述的信元数据处理装置, 其中, 所述信元数据为多 播信元数据时, 所述信元数据的目的标识为多播组标识;
所述多播路由表中多播组标识与下游设备的设备号对应。
11.如权利要求 10所述的信元数据处理装置, 其中, 所述交换网元还具 有与多播信元数据对应的多播负载均衡表, 所示多播负载均衡表中所述下游 设备的设备号与所述交换网元链路的链路号对应;
所述获取模块设置为: 根据所述多播信元数据的目的标识查询多播路由 表以及多播负载均衡表,获得与所述多播信元数据对应的输出端口号比特图。
12.如权利要求 8-11 任一项所述的信元数据处理装置, 其中, 所述负载 均衡处理模块包括:
第一计算模块, 设置为: 计算所述交换网元包括的多条链路中每条链路 的数据量累计值;
负载均衡处理子模块, 设置为: 根据所述数据量累计值, 得到所述负载 均衡结果。
13. 根据权利要求 12所述的信元数据处理装置, 其中, 所述目的端口号 确定模块包括:
选择模块, 设置为: 选择所述数据量累计值最小的链路所对应的端口号 作为所述输出目的端口号。
14.如权利要求 13所述的信元数据处理装置, 其中, 所述负载均衡处理 模块还包括:
判断模块, 设置为: 判断所述多条链路中的每条链路的数据量累计值是 否达到预设阔值, 获得判断结果;
第二计算模块, 设置为: 当判断结果指示所述多条链路中的任一链路的 数据量累计值达到预设阔值时, 获取所述数据量累计值达到阔值的链路的当 前緩存深度值;
重置模块, 设置为: 用所述数据量累计值达到预设阔值的链路的当前緩 存深度值重置所述链路的所述数据量累计值。
15.一种交换网元,所述交换网元具有与单播信元数据对应的单播路由表、 与多播信元数据对应的多播路由表、 以及与多播信元数据对应的多播负载均 衡表; 所述交换网元还包括权利要求 8-14任一项所述的信元数据处理装置。
PCT/CN2013/090991 2013-04-18 2013-12-31 一种信元数据处理方法、装置和交换网元 WO2014169691A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201310135750.5 2013-04-18
CN201310135750.5A CN104113486A (zh) 2013-04-18 2013-04-18 一种信元数据处理方法、装置和交换网元

Publications (1)

Publication Number Publication Date
WO2014169691A1 true WO2014169691A1 (zh) 2014-10-23

Family

ID=51710121

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2013/090991 WO2014169691A1 (zh) 2013-04-18 2013-12-31 一种信元数据处理方法、装置和交换网元

Country Status (2)

Country Link
CN (1) CN104113486A (zh)
WO (1) WO2014169691A1 (zh)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105515993A (zh) * 2015-11-26 2016-04-20 西安空间无线电技术研究所 一种光电混合交换内部路径映射方法

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111224888A (zh) * 2018-11-27 2020-06-02 华为技术有限公司 发送报文的方法及报文转发设备

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1758804A (zh) * 2004-10-10 2006-04-12 中兴通讯股份有限公司 一种时分组网系统中扇区负载均衡的方法
CN1874348A (zh) * 2006-01-10 2006-12-06 华为技术有限公司 一种多处理单元负载均衡方法和多处理单元系统
CN102281192A (zh) * 2011-07-19 2011-12-14 中兴通讯股份有限公司 交换网络芯片的信元处理方法及装置
CN102946617A (zh) * 2011-08-15 2013-02-27 中兴通讯股份有限公司 负荷分担方法、装置和系统

Family Cites Families (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US4621359A (en) * 1984-10-18 1986-11-04 Hughes Aircraft Company Load balancing for packet switching nodes
EP1168727B1 (en) * 2000-06-19 2006-05-03 Broadcom Corporation Memory management unit architecture for switch fabric
CN101119313A (zh) * 2007-09-14 2008-02-06 杭州华三通信技术有限公司 一种负载分担方法和设备
CN102111327B (zh) * 2009-12-29 2014-11-05 中兴通讯股份有限公司 信元调度方法和系统

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1758804A (zh) * 2004-10-10 2006-04-12 中兴通讯股份有限公司 一种时分组网系统中扇区负载均衡的方法
CN1874348A (zh) * 2006-01-10 2006-12-06 华为技术有限公司 一种多处理单元负载均衡方法和多处理单元系统
CN102281192A (zh) * 2011-07-19 2011-12-14 中兴通讯股份有限公司 交换网络芯片的信元处理方法及装置
CN102946617A (zh) * 2011-08-15 2013-02-27 中兴通讯股份有限公司 负荷分担方法、装置和系统

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105515993A (zh) * 2015-11-26 2016-04-20 西安空间无线电技术研究所 一种光电混合交换内部路径映射方法
CN105515993B (zh) * 2015-11-26 2019-02-15 西安空间无线电技术研究所 一种光电混合交换内部路径映射方法

Also Published As

Publication number Publication date
CN104113486A (zh) 2014-10-22

Similar Documents

Publication Publication Date Title
CN106998302B (zh) 一种业务流量的分配方法及装置
US11917262B2 (en) Dynamic designated forwarder election per multicast stream for EVPN all-active homing
US20130003549A1 (en) Resilient Hashing for Load Balancing of Traffic Flows
CN103095568B (zh) 机架式交换设备实现堆叠的系统及方法
CN104767694A (zh) 一种面向Fat-Tree数据中心网络架构的数据流转发方法
CN101873269B (zh) 数据转发设备和端口缓存的分配方法
CN103607358B (zh) 基于链路利用率平均和的动态ecmp 方法及系统
CN100553199C (zh) 基于pcie交换网的组播实现方法、系统及设备
CN113132249A (zh) 一种负载均衡方法和设备
CN103281257B (zh) 一种协议报文处理方法和设备
WO2014000399A1 (zh) 链路选择方法和装置
CN104106244A (zh) 控制装置、通信系统、通信方法和程序
CN113630347B (zh) 一种数据传输方法、系统、存储介质及设备
WO2013097639A1 (en) Expanding member ports of a link aggregation group between clusters
WO2013010396A1 (zh) 交换网络芯片的信元处理方法及装置
CN101483593A (zh) 一种交换设备中基于聚合链路分配缓存的方法及装置
CN104243348A (zh) 一种数据处理方法和装置
WO2014169691A1 (zh) 一种信元数据处理方法、装置和交换网元
CN103326916A (zh) 智能变电站自动划分并优化vlan的系统及方法
CN107276908B (zh) 一种路由信息处理方法及分组交换设备
US9154390B2 (en) Packet relay apparatus and measurement method for measuring discard number of data packets
CN101986616A (zh) 以太网分路器及其分路方法
CN105282036A (zh) 一种路由节点、路由交换方法及系统
CN104580000A (zh) 一种基于流表技术的RapidIO交换方法
US20150319080A1 (en) Basic self-routing unit and method for building its half-cleaners, sorters, network concentrators and multicast switching network

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: 13882572

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: 13882572

Country of ref document: EP

Kind code of ref document: A1