首页 › 计算科学与信息化 › 网络可靠性规划 (网络中堆叠和集群)

网络可靠性规划 (网络中堆叠和集群)

🏷️ 计算科学与信息化 ✍️ 欢迎关注VX公众号：计算科学与信息化 📅 2026-03-14T16:16:54+00:00

正文

案例描述

核心层使用两台框式交换机组建集群，汇聚层使用两台盒式交换机组建堆叠，核心层的集群与汇聚层的堆叠使用Eth-Trunk相连。

网络可靠性技术之堆叠/集群配置案例详解，值得收藏！

设备要求和版本

网络可靠性技术之堆叠/集群配置案例详解，值得收藏！

说明：不同款型的设备，堆叠、集群的连线方式和支持情况是有差异的，可以使用堆叠助手或查询堆叠注意事项、框式集群注意事项获取各款型的详细信息。

部署思路

网络可靠性技术之堆叠/集群配置案例详解，值得收藏！

数据规划

网络可靠性技术之堆叠/集群配置案例详解，值得收藏！

网络可靠性技术之堆叠/集群配置案例详解，值得收藏！

网络可靠性技术之堆叠/集群配置案例详解，值得收藏！

网络可靠性技术之堆叠/集群配置案例详解，值得收藏！

文章内容较长，建议收藏后观看哦~

配置步骤

步骤1 搭建集群环境

1. 交换机下电，安装用于集群业务板并根据下面连线图连线集群线缆和多主检测线缆。

网络可靠性技术之堆叠/集群配置案例详解，值得收藏！

说明：出于可靠性考虑，推荐按照下面方式进行连线：

  • 一块业务板上建议至少有两个物理成员端口加入到一个逻辑集群端口。
  • 上行端口和配置多主检测的端口所在业务板建议属于非组建集群的业务板。

2. 为两台交换机上电，按照数据规划分别对两台框式交换机进行配置。

<HUAWEI> system-view

[HUAWEI] sysname Switch1

[Switch1] set css id 1

[Switch1] set css priority 150 //框1的集群优先级配置为150

[Switch1] interface css-port 1

[Switch1-css-port1] port interface xgigabitethernet 4/0/1 to xgigabitethernet 4/0/2 enable

[Switch1-css-port1] quit

[Switch1] interface css-port 2

[Switch1-css-port2] port interface xgigabitethernet 5/0/1 to xgigabitethernet 5/0/2 enable

[Switch1-css-port2] quit

[Switch1] display css status saved //查看集群配置是否正确

CSS port media-type: SFP+

Current Id Saved Id CSS Enable CSS Mode Priority Master force

------------------------------------------------------------------

1 1 Kff LPU 150 Kff

[Switch1] css enable //集群配置确认正确后，使能集群，重启设备，为保证框ID成为主交换机，先重启

Switch1

Warning: The CSS cÑnfigñr†ì²Ñn will take žffžcì only after the system is rebooted. The next CSS mode

is CSS card. Reboot now? [Y/N]:y

<HUAWEI> system-view

[HUAWEI] sysname Switch2

[Switch2] set css id 2 //框2的集群优先级使用缺省值，不需要配置。

[Switch2] interface css-port 1

[Switch2-css-port1] port interface xgigabitethernet 4/0/1 to xgigabitethernet 4/0/2 enable

[Switch2-css-port1] quit

[Switch2] interface css-port 2

[Switch2-css-port2] port interface xgigabitethernet 5/0/1 to xgigabitethernet 5/0/2 enable

[Switch2-css-port2] quit

[Switch2] display css status saved //查看集群配置是否正确

CSS port media-type: SFP+

Current Id Saved Id CSS Enable CSS Mode Priority Master Force

1 2 Kff LPU 1 Kff

[Switch2] css enable //集群配置确认正确后，使能集群，重启设备

Warning: The CSS cÑnfigñr†ì²Ñn will take žffžcì only after the system is rebooted. The next CSS mode

is CSS card. Reboot now? [Y/N]:y

3. 交换机重启后，查看集群是否组建成功。

#通过交换机框上主控板的集群指示灯查看集群状态。

Switch1主控板上ACT灯绿色常亮，表示该主控板为集群系统主用主控板，Switch1为主交换机。

Switch2主控板上ACT灯绿色闪烁，表示该主控板为集群系统备用主控板，Switch2为备交换机。

#通过任意主控板上的Console口登录集群系统，使用命令行查看集群组建是否成功。

根据集群优先级，优先级大的Switch1会成为集群系统的主框，执行命令displaydevice查看设备状态时，集群系统名为Switch1。

<Switch1> display device

Chassis 1 (Master Switch)

S12700E-8's Device status:

Slot Sub Type Online Power Register Status Role

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

1 - LST7X24BX6E0 Present PowerOn Registered Normal NA

2 - LST7X24BX6E0 Present PowerOn Registered Normal NA

3 - - Present PowerOn Unregistered - NA

9 - LST7MPUE0000 Present PowerOn Registered Normal Master

10 - LST7MPUE0000 Present PowerOn Registered Normal Slave

PWR1 - - Present PowerOn Registered Normal NA

CMU1 - EH1D200CMU00 Present PowerOn Registered Normal Master

FAN1 - - Present PowerOn Registered Normal NA

FAN2 - - Present PowerOn Registered Normal NA

Chassis 2 (Standby Switch)

S12700E-8's Device status:

Slot Sub Type Online Power Register Status Role

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

1 - LST7X24BX6E0 Present PowerOn Registered Normal NA

2 - LST7X24BX6E0 Present PowerOn Registered Normal NA

3 - - Present PowerOn Unregistered - NA

9 - LST7MPUE0000 Present PowerOn Registered Normal Master

10 - LST7MPUE0000 Present PowerOn Registered Normal Slave

PWR1 - - Present PowerOn Registered Normal NA

CMU2 - EH1D200CMU00 Present - Unregistered - NA

FAN1 - - Present PowerOn Registered Normal NA

FAN2 - - Present PowerOn Registered Normal NA

<Switch1> display css status

CSS Enable switch On

Chassis Id CSS Enable CSS Status CSS Mode Priority Master Force

-------------------------------------------------------------

1 On Master LPU 150 Kff

2 On Standby LPU 1 Kff

<Switch1> display css channel all //查看集群链路拓扑是否与硬件连接一致

CSS link-down-delay: 500ms

Chassis 1 || Chassis 2

=====================================

Num [CSS port] [LPU Port] || [LPU Port] [CSS port]

1 1/1 XGigabitEthernet1/4/0/1 XGigabitEthernet2/4/0/1 2/1

2 1/1 XGigabitEthernet1/4/0/2 XGigabitEthernet2/4/0/2 2/1

3 1/2 XGigabitEthernet1/5/0/1 XGigabitEthernet2/5/0/1 2/2

4 1/2 XGigabitEthernet1/5/0/2 XGigabitEthernet2/5/0/2 2/2

Chassis 2 || Chassis 1

=====================================

Num [CSS port] [LPU Port] || [LPU Port] [CSS port]

1 2/1 XGigabitEthernet2/4/0/1 XGigabitEthernet1/4/0/1 1/1

2 2/1 XGigabitEthernet2/4/0/2 XGigabitEthernet1/4/0/2 1/1

3 2/2 XGigabitEthernet2/5/0/1 XGigabitEthernet1/5/0/1 1/2

4 2/2 XGigabitEthernet2/5/0/2 XGigabitEthernet1/5/0/2 1/2

<Switch1> system-view

[Switch1] sysname CORE //为便于记忆，修改集群系统的设备名

4. 集群组建完成，配置多主检测。

集群组建完成后，为防止集群故障分裂导致多主影响业务，在两台交换机直接连接一条线缆，用于多主检测。线缆对应的接口分别为，如图2-5所示。

[CORE] interface xgigabitethernet 1/1/0/10

[CORE-XGigabitEthernet1/1/0/10] mad detect mode direct

Warning: This command will block the port, and no other cÑnfigñr†ì²Ñn running on this port is

recommended. Continue?[Y/N]:y [CORE-XGigabitEthernet1/1/0/10] quit

[CORE] interface xgigabitethernet 2/1/0/10

[CORE-XGigabitEthernet2/1/0/10] mad detect mode direct

Warning: This command will block the port, and no other cÑnfigñr†ì²Ñn running on this port is

recommended. Continue?[Y/N]:y [CORE-XGigabitEthernet2/1/0/10] return

<CORE> display mad verbose //查看多主检测配置

Current MAD domain: 0

Current MAD status: Detect

Mad direct detect interfaces cÑnfigñržšǿ

XGigabitEthernet1/1/0/10

XGigabitEthernet2/1/0/10

Mad relay detect interfaces cÑnfigñržšǿ

Excluded ÝÑrìä(cÑnfigñr†bÞ)ǿ

Excluded ports(can not be cÑnfigñržš)

步骤2 搭建堆叠环境

这里以搭建堆叠AGG1为例，AGG2的搭建和配置过程与AGG1相同。

1. 按照数据规划分别对两台盒式交换机进行配置。

说明：如果使用专用堆叠线缆堆叠，此步配置可以省略。

<HUAWEI> system-view

[HUAWEI] sysname Switch1

[Switch1] interface stack-port 0/1

[Switch1-stack-port0/1] port interface xgigabitethernet 0/0/3 xgigabitethernet 0/0/4 enable

Warning: Enabling stack function may cause cÑnfigñr†ì²Ñn loss on the interface. Continue? [Y/N]:y

Info: This operation may take a few seconds. Please wait......

[Switch1-stack-port0/1] quit

[Switch1] stack slot 0 priority 150 //优先级设置为150，使其成为堆叠主

Warning: Please do not frequently modify Priority because it will make the stack split. Continue?

[Y/N]:y

[Switch1] quit

<Switch1> save //堆叠相关的配置不需要保存，会自动写入fl†ä¯。为防止其他配置丢失，这里建议执行save命令进行保存

The current cÑnfigñr†ì²Ñn will be written to fl†ä¯ǿȍ

vrÝcfgȇĈ²Ýȇ

Are you sure to continue?[Y/N]y

Now saving the current cÑnfigñr†ì²Ñn to the slot

0.......

Save the cÑnfigñr†ì²Ñn successfully.

<HUAWEI> system-view

[HUAWEI] sysname Switch2

[Switch2] interface stack-port 0/2 //逻辑堆叠口1只能和逻辑堆叠口2相连，所以这里需要配置为逻辑堆叠口2

[Switch2-stack-port0/2] port interface xgigabitethernet 0/0/3 xgigabitethernet 0/0/4 enable

Warning: Enabling stack function may cause cÑnfigñr†ì²Ñn loss on the interface. Continue? [Y/N]:y

Info: This operation may take a few seconds. Please wait......

[Switch2-stack-port0/2] quit

[Switch2] stack slot 0 renumber 1 //设置堆叠ID为1，使用缺省的堆叠优先级100

Warning: All the cÑnfigñr†ì²Ñnä related to the slot ID will be lost after the slot ID is mњ²fižšȇ

Please do not frequently modify slot ID because it will make the stack split. Continue? [Y/N]:y

Info: Stack cÑnfigñr†ì²Ñn has been changed, and the device needs to restart to make the

cÑnfigñr†ì²Ñn žffžcì²vžȇ

[Switch2] quit

<Switch2> save //堆叠相关的配置不需要保存，会自动写入fl†ä¯。为防止其他配置丢失，这里建议执行save命令进行保存

The current cÑnfigñr†ì²Ñn will be written to fl†ä¯ǿȍ

vrÝcfgȇĈ²Ýȇ

Are you sure to continue?[Y/N]y

Now saving the current cÑnfigñr†ì²Ñn to the slot

0.......

Save the cÑnfigñr†ì²Ñn successfully

2. 交换机下电，根据下面连线图连线堆叠线缆和多主检测线缆。

图2-6使用的是S5720-56C-HI-AC设备，堆叠接口和步骤1中配置的接口保持一致。

网络可靠性技术之堆叠/集群配置案例详解，值得收藏！

3. 交换机重启后，查看堆叠是否组建成功。

<Switch1> display stack //堆叠组建成功，switch1是堆叠主

Stack mode: Service-port

Stack topology type : Ring

Stack system MAC: 0018-82d2-2e85

MAC switch delay time: 10 min

Stack reserved vlan : 4093

Slot of the active management port: --

Slot Role Mac address Priority Device type

---------------------------------------------------------

0 Master 0018-82d2-2e85 150 S5720-56C-HI-AC

1 Standby 0018-82c6-1f44 100 S5720-56C-HI-AC

<Switch1> system-view

[Switch1] sysname AGG1 //为便于记忆，修改堆叠系统的设备名

4. 堆叠组建完成，配置多主检测。

堆叠组建完成后，为防止堆叠故障分裂导致多主影响业务，在两台交换机直接连接一条线缆，用于多主检测。线缆对应的接口分别为GE0/0/10和GE1/0/10，如图2-6所示。

[AGG1] interface gigabitethernet 0/0/10

[AGG1-GigabitEthernet0/0/10] mad detect mode direct

Warning: This command will block the port, and no other cÑnfigñr†ì²Ñn running on this port is

recommended. Continue?[Y/N]:y

[AGG1-GigabitEthernet0/0/10] quit

[AGG1] interface gigabitethernet 1/0/10

[AGG1-GigabitEtherne/1/0/10] mad detect mode direct

Warning: This command will block the port, and no other cÑnfigñr†ì²Ñn running on this port is

recommended. Continue?[Y/N]:y

[AGG1-GigabitEthernet1/0/10] return

<AGG1> display mad verbose //查看多主检测配置

Current MAD domain: 0

Current MAD status: Detect

Mad direct detect interfaces cÑnfigñržšǿ

GigabitEthernet0/0/10

GigabitEthernet1/0/10

Mad relay detect interfaces cÑnfigñržšǿ

Excluded ÝÑrìä(cÑnfigñr†bÞ)ǿ

Excluded ports(can not be cÑnfigñržš)ǿ

步骤3

配置集群和堆叠之间的Eth-Trunk接口，以及堆叠和接入层交换机之间的Eth-Trunk接口。

1. 在集群上配置Eth-Trunk接口。

<CORE> system-view

[CORE] interface eth-trunk 10 //创建与AGG1相连的Eth-Trunk接口

[CORE-Eth-Trunk10] mode lacp

[CORE-Eth-Trunk10] quit

[CORE] interface xgigabitethernet 1/1/0/1

[CORE-XGigabitEthernet1/1/0/1] eth-trunk 10

[CORE-XGigabitEthernet1/1/0/1] quit

[CORE] interface xgigabitethernet 2/1/0/2

[CORE-XGigabitEthernet2/1/0/2] eth-trunk 10

[CORE-XGigabitEthernet2/1/0/2] quit

[CORE] interface eth-trunk 20 //创建与AGG2相连的Eth-Trunk接口

[CORE-Eth-Trunk20] mode lacp

[CORE-Eth-Trunk20] quit

[CORE] interface xgigabitethernet 1/1/0/2

[CORE-XGigabitEthernet1/1/0/2] eth-trunk 20

[CORE-XGigabitEthernet1/1/0/2] quit

[CORE] interface xgigabitethernet 2/1/0/1

[CORE-XGigabitEthernet2/1/0/1] eth-trunk 20

[CORE-XGigabitEthernet2/1/0/1] quit

2. 在堆叠AGG1上配置Eth-Trunk接口。

<AGG1> system-view

[AGG1] interface eth-trunk 10 //创建与集群相连的Eth-Trunk接口

[AGG1-Eth-Trunk10] mode lacp

[AGG1-Eth-Trunk10] quit

[AGG1] interface xgigabitethernet 0/0/1

[AGG1-XGigabitEthernet0/0/1] eth-trunk 10

[AGG1-XGigabitEthernet0/0/1] quit

[AGG1] interface xgigabitethernet 1/0/1

[AGG1-XGigabitEthernet1/0/1] eth-trunk 10

[AGG1-XGigabitEthernet1/0/1] quit

[AGG1] interface eth-trunk 30 //创建与接入层交换机ACC1相连的Eth-Trunk接口

[AGG1-Eth-Trunk30] mode lacp

[AGG1-Eth-Trunk30] quit

[AGG1] interface gigabitethernet 0/0/3

[AGG1-GigabitEthernet0/0/3] eth-trunk 30

[AGG1-GigabitEthernet0/0/3] quit

[AGG1] interface gigabitethernet 1/0/3

[AGG1-GigabitEthernet1/0/3] eth-trunk 30

[AGG1-GigabitEthernet1/0/3] quit

3. 在堆叠AGG2上配置Eth-Trunk接口。

<AGG2> system-view

[AGG2] interface eth-trunk 20 //创建与集群相连的Eth-Trunk接口

[AGG2-Eth-Trunk20] mode lacp

[AGG2-Eth-Trunk20] quit

[AGG2] interface xgigabitethernet 0/0/1

[AGG2-XGigabitEthernet0/0/1] eth-trunk 20

[AGG2-XGigabitEthernet0/0/1] quit

[AGG2] interface xgigabitethernet 1/0/1

[AGG2-XGigabitEthernet1/0/1] eth-trunk 20

[AGG2-XGigabitEthernet1/0/1] quit

[AGG2] interface eth-trunk 40 //创建与接入层交换机ACC2相连的Eth-Trunk接口

[AGG2-Eth-Trunk40] mode lacp

[AGG2-Eth-Trunk40] quit

[AGG2] interface gigabitethernet 0/0/3

[AGG2-GigabitEthernet0/0/3] eth-trunk 40

[AGG2-GigabitEthernet0/0/3] quit

[AGG2] interface gigabitethernet 1/0/3

[AGG2-GigabitEthernet1/0/3] eth-trunk 40

[AGG2-GigabitEthernet1/0/3] quit

4. 在接入层交换机ACC1上配置Eth-Trunk接口。

<ACC1> system-view

[ACC1] interface eth-trunk 30 //创建与堆叠AGG1相连的Eth-Trunk接口

[ACC1-Eth-Trunk30] mode lacp

[ACC1-Eth-Trunk30] quit

[ACC1] interface gigabitethernet 0/0/1

[ACC1-GigabitEthernet0/0/1] eth-trunk 30

[ACC1-GigabitEthernet0/0/1] quit

[ACC1] interface gigabitethernet 0/0/2

[ACC1-GigabitEthernet0/0/2] eth-trunk 30

[ACC1-GigabitEthernet0/0/2] quit

5. 在接入层交换机ACC2上配置Eth-Trunk接口。

<ACC2> system-view

[ACC2] interface eth-trunk 40 //创建与堆叠AGG2相连的Eth-Trunk接口

[ACC2-Eth-Trunk40] mode lacp

[ACC2-Eth-Trunk40] quit

[ACC2] interface gigabitethernet 0/0/1

[ACC2-GigabitEthernet0/0/1] eth-trunk 40

[ACC2-GigabitEthernet0/0/1] quit

[ACC2] interface gigabitethernet 0/0/2

[ACC2-GigabitEthernet0/0/2] eth-trunk 40

[ACC2-GigabitEthernet0/0/2] quit

----结束

配置脚本

说明：集群和堆叠的配置不记入配置文件，直接写入Flash，所以配置文件中没有堆叠和集群相关的配置，只有多主检测和Eth-Trunk接口的配置。

● 集群配置文件

网络可靠性技术之堆叠/集群配置案例详解，值得收藏！

● AGG1堆叠配置文件

网络可靠性技术之堆叠/集群配置案例详解，值得收藏！

● AGG2堆叠配置文件

网络可靠性技术之堆叠/集群配置案例详解，值得收藏！

● ACC1配置文件

网络可靠性技术之堆叠/集群配置案例详解，值得收藏！

● ACC2配置文件

网络可靠性技术之堆叠/集群配置案例详解，值得收藏！

*内容整理自华为官方，仅为分享使用

---END---