WO2016091029A1 - 一种堆叠系统的流量转发的方法及装置 - Google Patents

一种堆叠系统的流量转发的方法及装置 Download PDF

Info

Publication number
WO2016091029A1
WO2016091029A1 PCT/CN2015/093232 CN2015093232W WO2016091029A1 WO 2016091029 A1 WO2016091029 A1 WO 2016091029A1 CN 2015093232 W CN2015093232 W CN 2015093232W WO 2016091029 A1 WO2016091029 A1 WO 2016091029A1
Authority
WO
WIPO (PCT)
Prior art keywords
port
physical port
available physical
member device
aggregation
Prior art date
Application number
PCT/CN2015/093232
Other languages
English (en)
French (fr)
Inventor
王树国
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 WO2016091029A1 publication Critical patent/WO2016091029A1/zh

Links

Images

Definitions

  • This document relates to, but is not limited to, the field of communication technologies, and in particular, to a method and apparatus for traffic forwarding of a stacking system.
  • a stacking system is a combination of multiple switch devices and virtualized into one device. This allows multiple devices in the stack to implement redundant backup between devices and link redundancy through cross-device link aggregation. Backup, improve the stability and reliability of the system, as shown in Figure 1, where the links contained in the ellipse line in the figure constitute a cross-device aggregation link.
  • the cross-device link aggregation technology can be used to configure a physical port on a member switch as a logical aggregate link port. When a member switch device fails or one link of a cross-device aggregation link fails, other interfaces can be used. A faulty switch device or link does not cause the entire aggregated link to fail, thus ensuring reliable transmission of data traffic.
  • the cross-device logical aggregation link port is implemented by creating an aggregated forwarding (Trunk) table on each device.
  • the trunk table contains all the physical ports that join the aggregated link, and the trunk table on each device is identical. As shown in Table 1, the trunk member represents a physical port.
  • the physical interfaces in the trunk table calculated according to the hash (HASH) algorithm will be located on different switches in the stack system. This may cause traffic that is forwarded from a logical aggregate link port to cross the device. Forwarding, which occupies the stacking bandwidth between stacked devices. Due to the limited bandwidth between stack devices, the traffic that is forwarded across devices increases the bandwidth load between stack devices and reduces the forwarding efficiency of traffic.
  • the HASH algorithm calculates the uncertainty of the port and makes it difficult for users to control the traffic forwarding of the cross-device link aggregation, which cannot meet the scenario of accurately controlling traffic forwarding.
  • the local traffic forwarding function is enabled.
  • the traffic that enters the device is preferentially forwarded from the egress port of the device.
  • the traffic will pass through the stacking bandwidth only after the link of the device fails.
  • the traffic indicated by the dotted line with the arrow is forwarded directly from the local device port and is not forwarded to the adjacent device across devices.
  • the local priority forwarding function only distinguishes the forwarding priority of the device and all other member devices.
  • the bandwidth of the stack can be reduced.
  • the traffic is re-enabled. Distributed on all other member devices. When the number of stack switches is large, traffic spans the stacking line between multiple devices, which greatly wastes limited stack bandwidth resources.
  • the embodiment of the invention provides a method and a device for forwarding traffic of a stacking system, so as to solve the technical problem of how to avoid the uncertainty of the traffic forwarding of the cross-device link aggregation and reduce the occupation of the stacking bandwidth.
  • An embodiment of the present invention provides a method for traffic forwarding of a stacking system, where the method includes:
  • the first member device of the stack system determines whether the outbound port of the data packet is an aggregate link port of the stack system.
  • the first physical port corresponding to the shortest path is selected according to the aggregation forwarding table maintained by the device. Multiple available physical ports, and distance values from the device to the path of the member device to which each available physical port belongs;
  • the data packet is forwarded through the first physical port.
  • the maintenance of the aggregation forwarding table includes:
  • the step of determining an available physical port corresponding to the aggregated link port includes:
  • the maintenance of the aggregation forwarding table further includes:
  • the available physical port corresponding to the aggregated link port is updated, and the aggregate forwarding table is generated and maintained according to the distance value of the path of the member device to the updated member device to which the available physical port belongs.
  • the maintenance of the aggregation forwarding table further includes: recording priority information corresponding to each available physical port;
  • the first physical port corresponding to the shortest path is selected from the first type of physical ports.
  • An embodiment of the present invention further provides a device for traffic forwarding of a stacking system, the device comprising:
  • the judging module is configured to determine, when the first member device of the stack system receives a data packet, whether the egress port of the data packet is an aggregation link port of the stack system;
  • the forwarding module is configured to forward the data packet through the first physical port.
  • the device further comprises:
  • the first maintenance module is configured to determine an available physical port corresponding to the aggregated link port, calculate a distance value of the path from the member device to the member device to which each available physical port belongs, and generate and maintain an aggregate forwarding table according to the distance value.
  • the maintenance module includes:
  • the first unit is configured to acquire first fault information of the member device detected by the stacking system
  • the second unit is configured to determine an available physical port corresponding to the aggregated link port according to the first fault information.
  • the device further comprises:
  • Obtaining a module configured to obtain second fault information of the physical port detected by the stacking system
  • the second maintenance module is configured to: according to the second fault information, update the available physical port corresponding to the aggregated link port, and generate and maintain the aggregation according to the distance value of the path of the member device to the updated member device of each available physical port. Forwarding table.
  • the device further comprises:
  • a recording module configured to record priority information corresponding to each available physical port
  • the selection module includes:
  • the third unit is configured to select the first type of physical port having the highest priority according to the priority information of the available physical port;
  • the fourth unit is configured to select a first physical port corresponding to the shortest path from the first type of physical ports.
  • the embodiment of the invention further provides a computer storage medium, wherein the computer storage medium stores computer executable instructions, and the computer executable instructions are used to execute the above method.
  • the distance value of the path between each member device in the stacking system and the priority information configured by the user for the member device are used in each
  • the aggregation forwarding table is set up in the member device.
  • the outbound port of the data packet received by the first member device in the stack system is an aggregated link port
  • the shortest path corresponding to the shortest path can be selected from the aggregation forwarding table of the device.
  • the first physical port forwards the data packet, which solves the uncertainty of the traffic forwarding of the inter-device link aggregation and reduces the occupation of the stack bandwidth.
  • FIG. 1 is a schematic diagram of networking of a stacking system in the related art
  • FIG. 2 is a schematic diagram of functions of local priority forwarding of a stacking system in related art
  • FIG. 3 is a flow chart of steps of a method for forwarding traffic of a stacking system according to an embodiment of the present invention
  • FIG. 4 is a schematic diagram of networking of three member device stacking systems according to an embodiment of the present invention.
  • FIG. 5 is a schematic structural diagram of an apparatus for performing traffic forwarding of a stacking system according to an embodiment of the present invention.
  • the embodiments of the present invention provide a method and a device for forwarding traffic of a stacking system, which can solve the problem of the aggregation of the traffic forwarding of the cross-device link in the related art. Uncertainty reduces the occupation of stacking bandwidth.
  • an embodiment of the present invention provides a method for traffic forwarding of a stacking system, where the method includes:
  • Step 31 After receiving the data packet, the first member device of the stack system determines whether the outbound port of the data packet is an aggregate link port of the stack system.
  • Step 32 When the outbound port of the data packet is an aggregated link port of the stack system, select the first physical port corresponding to the shortest path according to the aggregation forwarding table maintained by the device, where the aggregation forwarding table records the aggregation chain. The distance between the plurality of available physical ports corresponding to the path port and the path of the device to the member device to which each available physical port belongs;
  • step 33 the data packet is forwarded through the first physical port.
  • the aggregate link port of the stack system corresponds to each member device.
  • query the topology of the stack and the weight or priority of each member device According to the topology diagram of the stack system, and the device weight or priority, the shortest path algorithm (for example, Dijkstra algorithm, Floyd algorithm) is used to calculate the distance value of the path of any member device to each member device, and each calculated value is calculated.
  • the distance values and the corresponding physical ports are stored in each member device in the form of an aggregate forwarding table.
  • the outbound port of the data packet received by the first member device is an aggregated link port
  • the first physical port corresponding to the shortest path may be forwarded from the aggregate forwarding table of the first member device to forward the data packet.
  • three member devices form a networking example of a stacking system, in which A1, B1, B2, and C1 ports are configured to form a cross-device.
  • Logical aggregate link port Assuming that the distance between device A and B, B and C is 1, the trunk table settings of device A, device B, and device C are as shown in Table 2, Table 3, and Table 4, respectively.
  • the egress port of the traffic entering the device A is an aggregation link port, the actual physical egress port is calculated by the HASH algorithm on the physical port that is 0.
  • the egress port is A1 and the traffic is forwarded from the A1 port.
  • the trunk forwarding table settings of device A, device B, and device C are as shown in Table 5, Table 6, and Table 7, respectively.
  • the HASH algorithm calculates the port in the physical port with distance 1. The packets are forwarded through the B1 and B2 interfaces. That is, the traffic is directly forwarded through the device B. The device does not need to be forwarded from the device B to the device C. This effectively reduces the bandwidth occupation between the device B and the device C.
  • the maintenance of the aggregation forwarding table includes: determining an available physical port corresponding to the aggregated link port, and calculating a distance value of the path of the member device to the member device to which each available physical port belongs, according to the distance value. Generate and maintain an aggregate forwarding table.
  • each member device determines whether the configured physical port belongs to the same member device before generating and maintaining the respective aggregation forwarding table according to the distance value, and if the configured physical port is a cross-device aggregation chain
  • the port number of the member device is used to query the distance between the member device and the path of each member device.
  • the physical port of each member device is sorted according to the distance from small to large.
  • Each member device creates and fills in the aggregation forwarding table according to the sorted physical port sequence. If the configured physical port is not an aggregate link port across devices, the physical port is not sorted.
  • the step of determining an available physical port corresponding to the aggregated link port includes: acquiring first fault information of the member device detected by the stacking system; determining the aggregated link according to the first fault information The available physical port corresponding to the port.
  • the fault detection mechanism of the stack system detects whether a member device is faulty. If a member device is faulty, all member devices query whether there is a fault in the aggregation forwarding table. If the physical port of the member device is faulty, the physical port of the member device that is faulty is deleted, and the physical port corresponding to the aggregated link port is determined. Then, the member device is recalculated. The distance to the path of the member device to which each available physical port belongs, and sorts the remaining physical ports in descending order of distance values to generate and maintain a new aggregate forwarding table.
  • the maintenance of the aggregate forwarding table further includes: acquiring second fault information of the physical port detected by the stacking system; updating the available physical port corresponding to the aggregated link port according to the second fault information, and according to The member device generates and maintains an aggregate forwarding table from the distance value of the path of the member device to which each available physical port belongs.
  • the fault detection mechanism of the stack system detects whether a physical port is faulty. If a physical port fails, all member devices query whether there is a faulty physical port in the aggregation forwarding table. If there is a faulty physical port, the failed physical port is deleted from the aggregation forwarding table, and each remaining physical port is sorted according to the distance value from small to large, thereby generating and maintaining a new aggregation switch. Published.
  • the maintenance of the aggregation forwarding table further includes: recording priority information corresponding to each available physical port; and selecting, according to the aggregation forwarding table maintained by the device, the first physical port corresponding to the shortest path, including The first physical port corresponding to the shortest path is selected from the first type of physical port according to the priority information of the available physical port.
  • the user is configured to configure the priority information forwarded by the member device. If the priority information of the forwarding device is configured, the priority information of the user is configured according to the priority from highest to lowest. Each physical port is sorted. Of course, for a physical port that does not have a priority, the physical port is sorted according to the distance from the path to the new one. User-configured priority information forwards data packets. Of course, if the physical port of the first type of the highest priority has multiple physical ports, the first physical port corresponding to the shortest path can be selected from the first type of physical port to forward the data packet.
  • the above example is taken as an example.
  • the user configures the forwarding priority of each member device on the member device. The larger the value, the higher the priority. Assume that the priority of device A is set to 2 on device B, the priority of device B is 3, and the priority of device C is 1.
  • the Trunk table settings on device A, device B, and device C are shown in Table 8 and Table 9, respectively. And Table 10 is shown.
  • the trunk table on member devices A and C is still the trunk table calculated according to the shortest path.
  • the trunk table on device B is the trunk table sorted according to the priority and then sorted according to the shortest path.
  • the port is calculated from the high-priority physical port.
  • the port is B1 or B2, and traffic is forwarded from this port.
  • the Trunk table settings of Device A, Device B, and Device C are as shown in Table 11, Table 12, and Table 13, respectively.
  • the traffic entering Device B is from the device with priority 2.
  • the port is found to be A1, and the traffic is forwarded from A1, so that the purpose of accurately controlling traffic forwarding according to priority is achieved.
  • traffic forwarding on devices A and C is still forwarded according to the shortest path, which reduces the occupation of stack bandwidth.
  • Table 8 Trunk Table of Device A
  • Table 11 Trunk Table for Device A
  • P/L in Tables 2 to 13 above is the basis value for selecting the physical port, P is the priority, and L is the distance.
  • Table 2 Table 3, Table 4, Table 5, Table 6, Table 7, Table 8 P/L in Tables 10, 11 and 13 is the assumed distance value, and P/L in Table 9 and Table 12 are the values of the priority.
  • the embodiment of the invention further provides a computer storage medium, wherein the computer storage medium stores computer executable instructions, and the computer executable instructions are used to execute the above method.
  • an embodiment of the present invention further provides a device for traffic forwarding of a stacking system, where the device includes:
  • the determining module 51 is configured to determine, when the first member device of the stack system receives a data packet, whether the outbound port of the data packet is an aggregate link port of the stack system;
  • the selecting module 52 is configured to: when the egress port of the data packet is an aggregation link port of the stack system, select the first physical port corresponding to the shortest path according to the aggregation forwarding table maintained by the first member device, where the aggregation forwarding table The distance values of the plurality of available physical ports corresponding to the aggregated link port and the path of the device to the member device to which each available physical port belongs are recorded;
  • the forwarding module 53 is configured to forward the data packet through the first physical port.
  • the device further comprises:
  • the first maintenance module is configured to determine an available physical port corresponding to the aggregated link port, calculate a distance value of the path from the member device to the member device to which each available physical port belongs, and generate and maintain an aggregate forwarding table according to the distance value.
  • the first maintenance module includes:
  • the first unit is configured to acquire first fault information of the member device detected by the stacking system
  • the second unit is configured to determine an available physical port corresponding to the aggregated link port according to the first fault information.
  • the device further comprises:
  • Obtaining a module configured to obtain second fault information of the physical port detected by the stacking system
  • the second maintenance module is configured to: according to the second fault information, update the available physical port corresponding to the aggregated link port, and generate and maintain the aggregation according to the distance value of the path of the member device to the updated member device of each available physical port. Forwarding table.
  • the device further comprises:
  • a recording module configured to record priority information corresponding to each available physical port
  • the selection module 52 includes:
  • the third unit is configured to select the first type of physical port having the highest priority according to the priority information of the available physical port;
  • the fourth unit is configured to select a first physical port corresponding to the shortest path from the first type of physical ports.
  • all or part of the steps of the above embodiments may also be implemented by using an integrated circuit. These steps may be separately fabricated into individual integrated circuit modules, or multiple modules or steps may be fabricated into a single integrated circuit module. achieve.
  • the devices/function modules/functional units in the above embodiments may be implemented by a general-purpose computing device, which may be centralized on a single computing device or distributed over a network of multiple computing devices.
  • each device/function module/functional unit in the above embodiment When each device/function module/functional unit in the above embodiment is implemented in the form of a software function module and sold or used as a stand-alone product, it can be stored in a computer readable storage medium.
  • the above mentioned computer readable storage medium may be a read only memory, a magnetic disk or an optical disk or the like.
  • the foregoing technical solution solves the uncertainty of the traffic forwarding of the cross-device link aggregation and reduces the occupation of the stack bandwidth.

Landscapes

  • Data Exchanges In Wide-Area Networks (AREA)

Abstract

一种堆叠系统的流量转发的方法及装置，上述方法包括：堆叠系统的第一成员设备接收到一数据报文后，判断数据报文的出端口是否为堆叠系统的聚合链路端口；在数据报文的出端口为堆叠系统的聚合链路端口时，根据本设备维护的聚合转发表，选择出最短路径对应的第一物理端口，其中，聚合转发表中记录有聚合链路端口对应的多个可用物理端口、以及本设备到每个个可用物理端口所属成员设备的路径的距离值；将数据报文通过第一物理端口转发出去。上述技术方案能解决跨设备链路聚合流量转发的不确定性，减少对堆叠带宽的占用。

Description

一种堆叠系统的流量转发的方法及装置 技术领域
本文涉及但不限于通信技术领域，特别涉及一种堆叠系统的流量转发的方法及装置。
背景技术
堆叠系统是指把多台交换机设备组合在一起，虚拟化成一台设备，这样就可以通过堆叠系统内的多台设备实现设备间冗余备份，同时通过跨设备链路聚合技术实现链路冗余备份，提高了系统的稳定性和可靠性，如图1所示，其中，图中椭圆线包含的链路组成跨设备聚合链路。跨设备链路聚合技术可以将不同成员交换机上的物理端口配置为一个逻辑聚合链路端口，这样当某台成员交换机设备故障或者跨设备聚合链路其中一条链路出现故障时，可以使用其他的未故障交换机设备或链路，不会导致整个聚合链路失效，从而保证了数据流量的可靠传输。跨设备逻辑聚合链路端口在实现时是通过在每台设备上建立一张聚合转发(Trunk)表，Trunk表中包含加入聚合链路的所有物理端口，并且每个设备上Trunk表完全相同，如表1所示，其中Trunk Member表示物理端口。在进行流量转发时，根据哈希(HASH)算法计算得到的Trunk表中出物理接口会位于堆叠系统中的不同的交换机上，这样就可能导致从一个逻辑聚合链路端口转发的流量会跨设备转发，占用堆叠设备间堆叠带宽。由于堆叠设备间带宽有限，跨设备转发的流量增加了堆叠设备间带宽承载的压力，同时也降低了流量的转发效率。另外，HASH算法计算出端口的不确定性也让用户对于跨设备链路聚合流量转发的控制变的困难，无法满足需要精确控制流量转发的场景。
Figure PCTCN2015093232-appb-000001
表1
为了解决占用堆叠带宽的问题，出现了流量本地优先转发功能，即进入本设备的流量，优先从本设备的出端口转发出去，只有当本设备的链路出现故障后，流量才会通过堆叠带宽转发到其他成员设备上，如图2所示，带箭头的虚线所示流量直接从本设备端口转发出去，不会跨设备转发到相邻设备上。本地优先转发功能只是区分了本设备和其他所有成员设备的转发优先级，在设备和链路正常情况下是可以减少对堆叠带宽的占用，但是当设备或者链路出现故障后，流量又会重新分布在其他所有成员设备上，在堆叠交换机数量较多的时候，流量跨越多台设备间的堆叠线路，会极大的浪费有限的堆叠带宽资源。
发明内容
以下是对本文详细描述的主题的概述。本概述并非是为了限制权利要求的保护范围。
本发明实施例提供了一种堆叠系统的流量转发的方法及装置，以解决如何避免跨设备链路聚合流量转发的不确定性，实现减少对堆叠带宽的占用的技术问题。
本发明的实施例提供了一种堆叠系统的流量转发的方法，该方法包括：
堆叠系统的第一成员设备接收到一数据报文后，判断数据报文的出端口是否为堆叠系统的聚合链路端口；
在数据报文的出端口为堆叠系统的聚合链路端口时，根据本设备维护的聚合转发表，选择出最短路径对应的第一物理端口，其中，聚合转发表中记录有聚合链路端口对应的多个可用物理端口、以及本设备到每个可用物理端口所属成员设备的路径的距离值；
将数据报文通过第一物理端口转发出去。
其中，聚合转发表的维护包括：
确定聚合链路端口对应的可用物理端口，并计算本成员设备到每个可用物理端口所属成员设备的路径的距离值，根据距离值生成并维护聚合转发表。
其中，确定聚合链路端口对应的可用物理端口的步骤包括：
获取堆叠系统检测到的成员设备的第一故障信息；
根据第一故障信息确定聚合链路端口对应的可用物理端口。
其中，聚合转发表的维护还包括：
获取堆叠系统检测到的物理端口的第二故障信息；
根据第二故障信息，更新聚合链路端口对应的可用物理端口，并根据本成员设备到更新后的每个可用物理端口所属成员设备的路径的距离值生成并维护聚合转发表。
其中，
聚合转发表的维护还包括：记录每个可用物理端口对应的优先级信息；
根据本设备维护的聚合转发表，选择出最短路径对应的第一物理端口，包括：
按照可用物理端口的优先级信息，选择具有最高优先级的第一类物理端口；
从第一类物理端口中选择出最短路径对应的第一物理端口。
本发明的实施例还提供了一种堆叠系统的流量转发的装置，该装置包括：
判断模块，设置为当堆叠系统的第一成员设备接收到一数据报文后，判断数据报文的出端口是否为堆叠系统的聚合链路端口；
选择模块，设置为在数据报文的出端口为堆叠系统的聚合链路端口时，根据第一成员设备维护的聚合转发表，选择出最短路径对应的第一物理端口，其中，聚合转发表中记录有聚合链路端口对应的多个可用物理端口、以及本设备到每个可用物理端口所属成员设备的路径的距离值；
转发模块，设置为将数据报文通过第一物理端口转发出去。
其中，装置还包括：
第一维护模块，设置为确定聚合链路端口对应的可用物理端口，并计算本成员设备到每个可用物理端口所属成员设备的路径的距离值，根据距离值生成并维护聚合转发表。
其中，维护模块包括：
第一单元，设置为获取堆叠系统检测到的成员设备的第一故障信息；
第二单元，设置为根据第一故障信息确定聚合链路端口对应的可用物理端口。
其中，装置还包括：
获取模块，设置为获取堆叠系统检测到的物理端口的第二故障信息；
第二维护模块，设置为根据第二故障信息，更新聚合链路端口对应的可用物理端口，并根据本成员设备到更新后的每个可用物理端口所属成员设备的路径的距离值生成并维护聚合转发表。
其中，装置还包括：
记录模块，设置为记录每个可用物理端口对应的优先级信息；
选择模块包括：
第三单元，设置为按照可用物理端口的优先级信息，选择具有最高优先级的第一类物理端口；
第四单元，设置为从第一类物理端口中选择出最短路径对应的第一物理端口。
本发明实施例还提供了一种计算机存储介质，所述计算机存储介质中存储有计算机可执行指令，所述计算机可执行指令用于执行上述的方法。
本发明实施例的上述方案包括以下有益效果：
在本发明的实施例的堆叠系统的流量转发的方法中，通过堆叠系统中每个成员设备间的路径的距离值以及用户为成员设备配置的优先级信息，在每 个成员设备中建立聚合转发表，这样，当堆叠系统中的第一成员设备接收到的数据报文的出端口是聚合链路端口时，可以从本设备的聚合转发表中选择最短路径对应的第一物理端口将数据报文转发出去，从而解决了跨设备链路聚合流量转发的不确定性，减少对堆叠带宽的占用。
在阅读并理解了附图和详细描述后，可以明白其他方面。
附图概述
图1为相关技术中堆叠系统的组网示意图；
图2为相关技术中堆叠系统本地优先转发的功能示意图；
图3为本发明实施例中堆叠系统的流量转发的方法的步骤流程图；
图4为本发明实施例中三台成员设备堆叠系统的组网示意图；
图5为本发明实施例中堆叠系统的流量转发的装置的结构示意图。
本发明的实施方式
下面将结合附图及具体实施例进行详细描述。
本发明实施例针对相关技术中跨设备链路聚合流量转发的不确定性，同时浪费堆叠带宽的问题，提供了一种堆叠系统的流量转发的方法及装置，能解决跨设备链路聚合流量转发的不确定性，减少对堆叠带宽的占用。
如图3所示，在本发明的实施例提供了一种堆叠系统的流量转发的方法，该方法包括：
步骤31，堆叠系统的第一成员设备接收到一数据报文后，判断数据报文的出端口是否为堆叠系统的聚合链路端口；
步骤32，当数据报文的出端口为堆叠系统的聚合链路端口时，根据本设备维护的聚合转发表，选择出最短路径对应的第一物理端口，其中，聚合转发表中记录有聚合链路端口对应的多个可用物理端口、以及本设备到每个可用物理端口所属成员设备的路径的距离值；
步骤33，将数据报文通过第一物理端口转发出去。
在本发明的实施例中，堆叠系统的聚合链路端口对应每个成员设备的物 理端口，当堆叠系统设备初始化完成后，查询堆叠系统的拓扑结构图，以及每个成员设备配置的权重或者优先级。并根据堆叠系统的拓扑结构图，以及设备权重或者优先级，使用最短路径算法(例如Dijkstra算法、Floyd算法)计算任一成员设备到每个成员设备的路径的距离值，并将计算得到的每个距离值与对应的物理端口以聚合转发表的形式存储在每个成员设备中。这样当第一成员设备接收到的数据报文的出端口是聚合链路端口时，可以从第一成员设备的聚合转发表中选取最短路径对应的第一物理端口将该数据报文转发出去。
在本发明的实施例中，如图4所示，三台成员设备(A、B和C)组成堆叠系统的组网示例，在此堆叠系统上配置A1、B1、B2和C1端口组成跨设备逻辑聚合链路端口。假定设备A与B，B与C之间的距离都为1，那么设备A、设备B和设备C的Trunk表设置分别如表2、表3和表4所示。当进入设备A的流量的出端口为聚合链路端口时，首先在距离为0的物理端口中，通过HASH算法计算实际物理出端口，得到出端口为A1，流量从A1口转发出去。当链路A1出现故障后，设备A、设备B和设备C的Trunk转发表设置分别如表5、表6和表7所示，HASH算法在距离为1的物理端口中计算出端口，这样流量通过B1和B2口转发出去，即流量直接通过设备B转发出去，不需要从设备B转发到设备C，有效的降低了设备B和设备C之间的堆叠带宽占用。
P/L Member
0 A1
1 B1
1 B2
2 C1
表2：设备A的Trunk表
P/L Member
0 B1
0 B2
1 A1
1 C1
表3：设备B的Trunk表
P/L Member
0 C1
1 B1
1 B2
2 A1
表4：设备C的Trunk表
P/L Member
1 B1
1 B2
2 C1
表5：设备A的Trunk表
P/L Member
0 B1
0 B2
1 C1
表6：设备B的Trunk表
P/L Member
0 C1
1 B1
1 B2
表7：设备C的Trunk表
在本发明的上述实施例中，聚合转发表的维护包括：确定聚合链路端口对应的可用物理端口，并计算本成员设备到每个可用物理端口所属成员设备的路径的距离值，根据距离值生成并维护聚合转发表。
在本发明的实施例中，每个成员设备在根据距离值生成并维护各自的聚合转发表之前，会先判断配置的物理端口是否属于同一成员设备，若配置的物理端口是跨设备的聚合链路端口，则以具体物理端口为关键字查询本设备存储的本成员设备到每个成员设备的路径的距离值，并按照距离值从小到大的方式对每个成员设备的物理端口进行排序，每个成员设备按照排序后的物理端口顺序创建并填写聚合转发表，若配置的物理端口不是跨设备的聚合链路端口，则不对该物理端口进行排序。
可选地，在本发明的上述实施例中，确定聚合链路端口对应的可用物理端口的步骤包括：获取堆叠系统检测到的成员设备的第一故障信息；根据第一故障信息确定聚合链路端口对应的可用物理端口。
在本发明的实施例中，堆叠系统的自身故障检测机制会检测是否有成员设备出现故障，若检测出某个成员设备出现故障，则所有成员设备查询自己的聚合转发表中是否存在出现故障的成员设备的物理端口，若存在出现故障的成员设备的物理端口，则把出现故障的成员设备的所有物理端口都删除，从而确定出聚合链路端口对应的可用物理端口，再重新计算本成员设备到每个可用物理端口所属成员设备的路径的距离值，并按照距离值从小到大的顺序对剩余物理端口进行排序，从而生成并维护新的聚合转发表。
在本发明的上述实施例中，聚合转发表的维护还包括：获取堆叠系统检测到的物理端口的第二故障信息；根据第二故障信息，更新聚合链路端口对应的可用物理端口，并根据本成员设备到更新后的每个可用物理端口所属成员设备的路径的距离值生成并维护聚合转发表。
在本发明的实施例中，堆叠系统的自身故障检测机制会检测是否有物理端口出现故障，若有物理端口出现故障，则所有成员设备查询自己的聚合转发表中是否存在出现故障的物理端口，若有出现故障的物理端口，则将该出现故障的物理端口从聚合转发表中删除，并按照距离值从小到大的顺序对剩余的每个物理端口进行排序，从而生成并维护新的聚合转发表。
在本发明的上述实施例中，聚合转发表的维护还包括：记录每个可用物理端口对应的优先级信息；根据本设备维护的聚合转发表，选择出最短路径对应的第一物理端口，包括：按照可用物理端口的优先级信息，选择具有最高优先级的第一类物理端口；从第一类物理端口中选择出最短路径对应的第一物理端口。
在本发明的实施例中，会检测用户是否配置成员设备转发的优先级信息，若配置了转发的优先级信息，便会根据用户配置的优先级信息，按照优先级从高到低的顺序对每个可用物理端口进行排序，当然，对于没有配置优先级的物理端口，会按照上述路径的距离值从小到大的顺序对物理端口进行排序，从而得到新的聚合转发表，使得成员设备可以按照用户配置的优先级信息将数据报文转发出去。当然，如果最高优先级的第一类物理端口有多个物理端口的话，可以从第一类物理端口中选择最短路径对应的第一物理端口来转发数据报文。
在本发明的实施例中，还以上述实例为例说明，为了精确控制转发流量时，用户在成员设备上配置每个成员设备的转发优先级，数值越大，优先级越高。假定用户在设备B上配置设备A的优先级为2，设备B的优先级为3，设备C优先级为1，那么设备A、设备B和设备C上Trunk表设置分别如表8、表9和表10所示。成员设备A和C上的Trunk表依然为前面按照最短路径计算得出的Trunk表，设备B上的Trunk表为先按照优先级排序，再按照最短路径排序得出的Trunk表。这样当进入设备B的流量的出端口为聚合链 路端口时，首先从高优先级物理端口中计算出端口，得到出端口为B1或者B2，流量从此端口转发出去。当B1和B2链路发生故障后，设备A、设备B和设备C的Trunk表设置分别如表11、表12和表13所示，这时进入设备B的流量从优先级为2的设备上查找出端口得到A1，流量从A1转发出去，从而达到了按照优先级精确控制流量转发的目的。同时设备A和C上的流量转发依然按照最短路径转发，减少了对堆叠带宽的占用。
P/L Member
0 A1
1 B1
1 B2
2 C1
表8：设备A的Trunk表
P/L Member
3 B1
3 B2
2 A1
1 C1
表9：设备B的Trunk表
P/L Member
0 C1
1 B1
1 B2
2 A1
表10：设备C的Trunk表
P/L Member
0 A1
2 C1
表11：设备A的Trunk表
P/L Member
2 A1
1 C1
表12：设备B的Trunk表
P/L Member
0 C1
2 A1
表13：设备C的Trunk表
其中，上述表2～表13中的P/L为选择物理端口的依据值，P表示优先级，L表示距离，表2、表3、表4、表5、表6、表7、表8、表10、表11和表13中的P/L为假定的距离值，表9和表12中的P/L为优先级的数值。
本发明实施例还提供了一种计算机存储介质，所述计算机存储介质中存储有计算机可执行指令，所述计算机可执行指令用于执行上述的方法。
为了更好的实现上述目的，如图5所示，本发明的实施例还提供了一种堆叠系统的流量转发的装置，该装置包括：
判断模块51，设置为当堆叠系统的第一成员设备接收到一数据报文后，判断数据报文的出端口是否为堆叠系统的聚合链路端口；
选择模块52，设置为在数据报文的出端口为堆叠系统的聚合链路端口时，根据第一成员设备维护的聚合转发表，选择出最短路径对应的第一物理端口，其中，聚合转发表中记录有聚合链路端口对应的多个可用物理端口、以及本设备到每个可用物理端口所属成员设备的路径的距离值；
转发模块53，设置为将数据报文通过第一物理端口转发出去。
其中，装置还包括：
第一维护模块，设置为确定聚合链路端口对应的可用物理端口，并计算本成员设备到每个可用物理端口所属成员设备的路径的距离值，根据距离值生成并维护聚合转发表。
其中，第一维护模块包括：
第一单元，设置为获取堆叠系统检测到的成员设备的第一故障信息；
第二单元，设置为根据第一故障信息确定聚合链路端口对应的可用物理端口。
其中，装置还包括：
获取模块，设置为获取堆叠系统检测到的物理端口的第二故障信息；
第二维护模块，设置为根据第二故障信息，更新聚合链路端口对应的可用物理端口，并根据本成员设备到更新后的每个可用物理端口所属成员设备的路径的距离值生成并维护聚合转发表。
其中，装置还包括：
记录模块，设置为记录每个可用物理端口对应的优先级信息；
此时，选择模块52包括：
第三单元，设置为按照可用物理端口的优先级信息，选择具有最高优先级的第一类物理端口；
第四单元，设置为从第一类物理端口中选择出最短路径对应的第一物理端口。
以上内容是结合具体的实施方式对本发明所作的进一步详细说明，不能认定本发明的具体实施只局限于这些说明。
本领域普通技术人员可以理解上述实施例的全部或部分步骤可以使用计算机程序流程来实现，所述计算机程序可以存储于一计算机可读存储介质中，所述计算机程序在相应的硬件平台上(如系统、设备、装置、器件等)执行，在执行时，包括方法实施例的步骤之一或其组合。
可选地，上述实施例的全部或部分步骤也可以使用集成电路来实现，这些步骤可以被分别制作成一个个集成电路模块，或者将它们中的多个模块或步骤制作成单个集成电路模块来实现。
上述实施例中的各装置/功能模块/功能单元可以采用通用的计算装置来实现，它们可以集中在单个的计算装置上，也可以分布在多个计算装置所组成的网络上。
上述实施例中的各装置/功能模块/功能单元以软件功能模块的形式实现并作为独立的产品销售或使用时，可以存储在一个计算机可读取存储介质中。上述提到的计算机可读取存储介质可以是只读存储器，磁盘或光盘等。
工业实用性
上述技术方案解决了跨设备链路聚合流量转发的不确定性，减少对堆叠带宽的占用。

Claims (11)

  1. 一种堆叠系统的流量转发的方法，包括：
    所述堆叠系统的第一成员设备接收到一数据报文后，判断所述数据报文的出端口是否为所述堆叠系统的聚合链路端口；
    在所述数据报文的出端口为所述堆叠系统的聚合链路端口时，根据本设备维护的聚合转发表，选择出最短路径对应的第一物理端口，其中，所述聚合转发表中记录有所述聚合链路端口对应的多个可用物理端口、以及本设备到每个可用物理端口所属成员设备的路径的距离值；
    将所述数据报文通过所述第一物理端口转发出去。
  2. 如权利要求1所述的方法，其中，所述聚合转发表的维护包括：
    确定所述聚合链路端口对应的可用物理端口，并计算本成员设备到每个所述可用物理端口所属成员设备的路径的距离值，根据所述距离值生成并维护所述聚合转发表。
  3. 如权利要求2所述的方法，其中，所述确定所述聚合链路端口对应的可用物理端口的步骤包括：
    获取堆叠系统检测到的成员设备的第一故障信息；
    根据所述第一故障信息确定所述聚合链路端口对应的可用物理端口。
  4. 如权利要求2所述的方法，所述聚合转发表的维护还包括：
    获取堆叠系统检测到的物理端口的第二故障信息；
    根据所述第二故障信息，更新所述聚合链路端口对应的可用物理端口，并根据本成员设备到更新后的每个可用物理端口所属成员设备的路径的距离值生成并维护所述聚合转发表。
  5. 如权利要求1所述的方法，其中，所述聚合转发表的维护包括：
    记录每个可用物理端口对应的优先级信息；
    所述根据本设备维护的聚合转发表，选择出最短路径对应的第一物理端口，包括：
    按照可用物理端口的优先级信息，选择具有最高优先级的第一类物理端口；
    从所述第一类物理端口中选择出最短路径对应的第一物理端口。
  6. 一种堆叠系统的流量转发的装置，包括：
    判断模块，设置为当所述堆叠系统的第一成员设备接收到一数据报文后，判断所述数据报文的出端口是否为所述堆叠系统的聚合链路端口；
    选择模块，设置为在所述数据报文的出端口为所述堆叠系统的聚合链路端口时，根据第一成员设备维护的聚合转发表，选择出最短路径对应的第一物理端口，其中，所述聚合转发表中记录有所述聚合链路端口对应的多个可用物理端口、以及本设备到每个可用物理端口所属成员设备的路径的距离值；
    转发模块，设置为将所述数据报文通过所述第一物理端口转发出去。
  7. 如权利要求6所述的装置，所述装置还包括：
    第一维护模块，设置为确定所述聚合链路端口对应的可用物理端口，并计算本成员设备到每个所述可用物理端口所属成员设备的路径的距离值，根据所述距离值生成并维护所述聚合转发表。
  8. 如权利要求7所述的装置，其中，所述第一维护模块包括：
    第一单元，设置为获取堆叠系统检测到的成员设备的第一故障信息；
    第二单元，设置为根据所述第一故障信息确定所述聚合链路端口对应的可用物理端口。
  9. 如权利要求7所述的装置，所述装置还包括：
    获取模块，设置为获取堆叠系统检测到的物理端口的第二故障信息；
    第二维护模块，设置为根据所述第二故障信息，更新所述聚合链路端口对应的可用物理端口，并根据本成员设备到更新后的每个可用物理端口所属成员设备的路径的距离值生成并维护所述聚合转发表。
  10. 如权利要求6所述的装置，所述装置还包括：
    记录模块，设置为记录每个可用物理端口对应的优先级信息；
    所述选择模块包括：
    第三单元，设置为按照可用物理端口的优先级信息，选择具有最高优先级的第一类物理端口；
    第四单元，设置为从所述第一类物理端口中选择出最短路径对应的第一 物理端口。
  11. 一种计算机存储介质，所述计算机存储介质中存储有计算机可执行指令，所述计算机可执行指令用于执行权利要求1～5中任一项所述的方法。
PCT/CN2015/093232 2014-12-08 2015-10-29 一种堆叠系统的流量转发的方法及装置 WO2016091029A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201410741242.6 2014-12-08
CN201410741242.6A CN105743801A (zh) 2014-12-08 2014-12-08 一种堆叠系统的流量转发的方法及装置

Publications (1)

Publication Number Publication Date
WO2016091029A1 true WO2016091029A1 (zh) 2016-06-16

Family

ID=56106657

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2015/093232 WO2016091029A1 (zh) 2014-12-08 2015-10-29 一种堆叠系统的流量转发的方法及装置

Country Status (2)

Country Link
CN (1) CN105743801A (zh)
WO (1) WO2016091029A1 (zh)

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US10007870B2 (en) * 2015-03-27 2018-06-26 Kyocera Document Solutions Inc. Redirection of a file based on distance between physical ports of an intermediate device
CN113472648A (zh) * 2021-06-18 2021-10-01 新华三信息安全技术有限公司 一种双向转发检测bfd方法及装置、网络设备
CN114374659A (zh) * 2022-01-18 2022-04-19 苏州盛科科技有限公司 一种堆叠系统中实现本地优先转发的方法、装置及存储介质
CN114760245A (zh) * 2022-03-28 2022-07-15 新华三技术有限公司合肥分公司 一种堆叠成员设备及其转发广播报文的方法

Families Citing this family (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107547243B (zh) * 2017-05-24 2021-04-02 新华三技术有限公司 一种报文转发方法及装置
CN108199986B (zh) * 2017-12-15 2022-05-27 瑞斯康达科技发展股份有限公司 一种数据传输方法、堆叠设备及堆叠系统
CN108600100B (zh) * 2018-07-13 2021-09-07 新华三技术有限公司合肥分公司 成员端口选取方法和装置
CN109039424B (zh) * 2018-07-18 2020-06-12 北京邮电大学 卫星星间网络通信路径确定方法、装置及电子设备
CN111224803B (zh) * 2018-11-23 2023-08-04 中兴通讯股份有限公司 一种堆叠系统中多主检测方法及堆叠系统
CN112087376B (zh) * 2019-06-14 2023-03-14 中兴通讯股份有限公司 一种负载分担的方法及装置
CN111030937A (zh) * 2019-12-16 2020-04-17 迈普通信技术股份有限公司 一种报文的转发方法、装置及存储介质
CN111740917A (zh) * 2020-03-23 2020-10-02 北京京东尚科信息技术有限公司 一种报文转发方法和装置
CN114513478A (zh) * 2022-01-07 2022-05-17 锐捷网络股份有限公司 堆叠系统的防丢包方法及装置

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101340456A (zh) * 2008-08-15 2009-01-07 杭州华三通信技术有限公司 一种分布式链路聚合故障收敛方法和一种堆叠设备
CN102780629A (zh) * 2012-07-31 2012-11-14 杭州华三通信技术有限公司 数据中心网络中的报文转发方法和装置
CN103780480A (zh) * 2012-10-17 2014-05-07 杭州华三通信技术有限公司 报文转发方法和设备
US20140140199A1 (en) * 2012-11-16 2014-05-22 Brocade Communications Systems, Inc. Virtual link aggregations across multiple fabric switches

Family Cites Families (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101599909B (zh) * 2009-07-08 2012-02-08 杭州华三通信技术有限公司 一种流量转发的方法和装置
CN101651680A (zh) * 2009-09-14 2010-02-17 杭州华三通信技术有限公司 一种网络安全部署方法和一种网络安全设备

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101340456A (zh) * 2008-08-15 2009-01-07 杭州华三通信技术有限公司 一种分布式链路聚合故障收敛方法和一种堆叠设备
CN102780629A (zh) * 2012-07-31 2012-11-14 杭州华三通信技术有限公司 数据中心网络中的报文转发方法和装置
CN103780480A (zh) * 2012-10-17 2014-05-07 杭州华三通信技术有限公司 报文转发方法和设备
US20140140199A1 (en) * 2012-11-16 2014-05-22 Brocade Communications Systems, Inc. Virtual link aggregations across multiple fabric switches

Cited By (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US10007870B2 (en) * 2015-03-27 2018-06-26 Kyocera Document Solutions Inc. Redirection of a file based on distance between physical ports of an intermediate device
CN113472648A (zh) * 2021-06-18 2021-10-01 新华三信息安全技术有限公司 一种双向转发检测bfd方法及装置、网络设备
CN113472648B (zh) * 2021-06-18 2023-10-24 新华三信息安全技术有限公司 一种双向转发检测bfd方法及装置、网络设备
CN114374659A (zh) * 2022-01-18 2022-04-19 苏州盛科科技有限公司 一种堆叠系统中实现本地优先转发的方法、装置及存储介质
CN114760245A (zh) * 2022-03-28 2022-07-15 新华三技术有限公司合肥分公司 一种堆叠成员设备及其转发广播报文的方法
CN114760245B (zh) * 2022-03-28 2024-02-09 新华三技术有限公司合肥分公司 一种堆叠成员设备及其转发广播报文的方法

Also Published As

Publication number Publication date
CN105743801A (zh) 2016-07-06

Similar Documents

Publication Publication Date Title
WO2016091029A1 (zh) 一种堆叠系统的流量转发的方法及装置
US10164782B2 (en) Method and system for constructing a loop free multicast tree in a data-center fabric
US8995277B2 (en) Method for dynamic load balancing of network flows on LAG interfaces
US9083645B2 (en) Systems and methods providing reverse path forwarding compliance for a multihoming virtual routing bridge
EP2883334B1 (en) Techniques for flooding optimization for link state protocols in a network topology
EP2928131B1 (en) Systems and methods for multipath load balancing
EP3232607B1 (en) Method and apparatus for establishing multicast group in fat-tree network
US10367686B2 (en) Automatically detecting roles of nodes in layered network topologies
JP2013510459A (ja) 分離的なパス計算アルゴリズム
CN104954276B (zh) 用于负载平衡多播流量的系统和方法
EP3213441B1 (en) Redundancy for port extender chains
US11818038B2 (en) Initiator-based data-plane validation for segment routed, multiprotocol label switched (MPLS) networks
WO2016153506A1 (en) Fast failover recovery in software defined networks
US10284457B2 (en) System and method for virtual link trunking
US10033666B2 (en) Techniques for virtual Ethernet switching of a multi-node fabric
US9705740B2 (en) Using unified API to program both servers and fabric for forwarding for fine-grained network optimizations
US11695856B2 (en) Scheduling solution configuration method and apparatus, computer readable storage medium thereof, and computer device
US8634417B2 (en) Method and apparatus providing selective flow redistribution across Multi Link Trunk/Link Aggregation Group (MLT/LAG) after port member failure and recovery
US20130070761A1 (en) Systems and methods for controlling a network switch
US10797991B2 (en) Setting link aggregation group
CN108390780B (zh) 用于处理信息的方法和装置
WO2024001315A1 (zh) 网元切换方法、装置、跨设备链路聚合组及存储介质

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: 15867288

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: 15867288

Country of ref document: EP

Kind code of ref document: A1