06-M-LAG+RDMA配置举例
本章节下载 (600.71 KB)
如图1-1所示,在数据中心(DC)内部构建一个稳定的二层网络,实现跨Leaf和跨Spine流量互通:
· Leaf作为接入交换机,部署M-LAG提高可靠性。对于M-LAG链路,进行负载分担提高链路利用率。Leaf配置VLAN双活网关或者VRRP,以满足服务器Server的入网需求。
· Spine作为汇聚层设备,部署RR,负责在各个Leaf间反射BGP路由。
· 在Leaf和Spine上部署PFC、数据缓冲区等智能无损网络特性,使以太网具备智能化拥塞控制功能,为RDMA提供零丢包的网络环境。本文以实现802.1p优先级为3的报文的无损转发为例。
· Leaf 配置DHCP Relay来转发DHCP报文,协助服务器获取IP地址。
设备 |
接口 |
IP地址 |
对接设备及接口 |
Leaf 1 |
WGE1/0/1 |
M-LAG成员端口 |
服务器Server1 |
WGE1/0/2 |
M-LAG成员端口 |
服务器Server2 |
|
HGE1/0/29 |
peer-link链路成员端口 |
Leaf 2:HGE1/0/29 |
|
HGE1/0/30 |
peer-link链路成员端口 |
Leaf 2:HGE1/0/30 |
|
HGE1/0/31 |
peer-link链路成员端口 |
Leaf 2:HGE1/0/31 |
|
HGE1/0/32 |
peer-link链路成员端口 |
Leaf 2:HGE1/0/32 |
|
WGE1/0/55 |
RAGG 1000 1.1.1.1/30 |
Keepalive口-Leaf 2:WGE1/0/55 |
|
WGE1/0/56 |
Keepalive口-Leaf 2:WGE1/0/56 |
||
HGE1/0/25 |
172.16.2.154/30 |
Spine 1:HGE1/1/1 |
|
HGE1/0/26 |
172.16.3.154/30 |
Spine 2:HGE1/1/1 |
|
Loopback1 |
50.50.255.41/32 |
BGP route-id |
|
Leaf 2 |
WGE1/0/1 |
M-LAG成员端口 |
服务器Server1 |
WGE1/0/2 |
M-LAG成员端口 |
服务器Server2 |
|
HGE1/0/29 |
peer-link链路成员端口 |
Leaf 1:HGE1/0/29 |
|
HGE1/0/30 |
peer-link链路成员端口 |
Leaf 1:HGE1/0/30 |
|
HGE1/0/31 |
peer-link链路成员端口 |
Leaf 1:HGE1/0/31 |
|
HGE1/0/32 |
peer-link链路成员端口 |
Leaf 1:HGE1/0/32 |
|
WGE1/0/55 |
RAGG 1000 1.1.1.2/30 |
Keepalive口-Leaf 1:WGE1/0/55 |
|
WGE1/0/56 |
Keepalive口-Leaf 1:WGE1/0/56 |
||
HGE1/0/25 |
172.16.2.158/30 |
Spine 1:HGE1/1/2 |
|
HGE1/0/26 |
172.16.3.158/30 |
Spine 2:HGE1/1/2 |
|
Loopback1 |
50.50.255.42/32 |
BGP route-id |
|
Leaf 3 |
WGE1/0/1 |
单挂接口 |
服务器Server3 |
WGE1/0/2 |
单挂接口 |
服务器Server4 |
|
HGE1/0/29 |
peer-link链路成员端口 |
Leaf 4:HGE1/0/29 |
|
HGE1/0/30 |
peer-link链路成员端口 |
Leaf 4:HGE1/0/30 |
|
HGE1/0/31 |
peer-link链路成员端口 |
Leaf 4:HGE1/0/31 |
|
HGE1/0/32 |
peer-link链路成员端口 |
Leaf 4:HGE1/0/32 |
|
WGE1/0/55 |
RAGG 1000 1.1.1.1/30 |
Keepalive口-Leaf 1:WGE1/0/55 |
|
WGE1/0/56 |
Keepalive口-Leaf 1:WGE1/0/56 |
||
HGE1/0/25 |
172.16.2.82/30 |
Spine 1:HGE1/1/3 |
|
HGE1/0/26 |
172.16.3.82/30 |
Spine 2:HGE1/1/3 |
|
Loopback1 |
50.50.255.23/32 |
BGP route-id |
|
Leaf 4 |
WGE1/0/1 |
单挂接口 |
服务器Server3 |
WGE1/0/2 |
单挂接口 |
服务器Server4 |
|
HGE1/0/29 |
peer-link链路成员端口 |
Leaf 3:HGE1/0/29 |
|
HGE1/0/30 |
peer-link链路成员端口 |
Leaf 3:HGE1/0/30 |
|
HGE1/0/29 |
peer-link链路成员端口 |
Leaf 3:HGE1/0/31 |
|
HGE1/0/30 |
peer-link链路成员端口 |
Leaf 3:HGE1/0/32 |
|
WGE1/0/55 |
RAGG 1000 1.1.1.2/30 |
Keepalive口-Leaf 1:WGE1/0/55 |
|
WGE1/0/56 |
Keepalive口-Leaf 1:WGE1/0/56 |
||
HGE1/0/25 |
172.16.2.86/30 |
Spine 1:HGE1/1/4 |
|
HGE1/0/26 |
172.16.3.86/30 |
Spine 2:HGE1/1/4 |
|
Loopback1 |
50.50.255.24/32 |
BGP route-id |
|
Spine 1 |
HGE1/1/1 |
172.16.2.153/30 |
Leaf 1:HGE1/0/25 |
HGE1/1/2 |
172.16.2.157/30 |
Leaf 2:HGE1/0/25 |
|
HGE1/1/3 |
172.16.2.81/30 |
Leaf 3:HGE1/0/25 |
|
HGE1/1/4 |
172.16.2.85/30 |
Leaf 4:HGE1/0/25 |
|
Loopback1 |
50.50.255.1/32 |
BGP route-id |
|
Spine 2 |
HGE1/1/1 |
172.16.3.153/30 |
Leaf 1:HGE1/0/26 |
HGE1/1/2 |
172.16.3.157/30 |
Leaf 2:HGE1/0/26 |
|
HGE1/1/3 |
172.16.3.81/30 |
Leaf 3:HGE1/0/26 |
|
HGE1/1/4 |
172.16.3.85/30 |
Leaf 4:HGE1/0/26 |
|
Loopback1 |
50.50.255.2/32 |
BGP route-id |
流量互通模型有以下几种:
· 同组M-LAG Leaf 服务器与服务器之间二层互通;
· 同组M-LAG Leaf 服务器与服务器之间三层互通;
· 不同M-LAG Leaf 服务器与服务器之间三层互通;
角色 |
设备 |
软件版本 |
Spine |
S12500X-AF/S12500F-AF |
不推荐使用 |
S12500R |
不推荐使用 |
|
S12500G-AF |
不推荐使用 |
|
S9820-8C(本文以S9820-8C作Spine为例) |
R6710版本 |
|
Leaf |
S6800/S6860/S6900 |
不推荐使用 |
S6812/S6813 |
不推荐使用 |
|
S6805/S6825/S6850/S9850(本文以S6850 作Leaf为例) |
R6710版本 |
|
S6890 |
不推荐使用 |
|
S9820-64H |
R6710版本 |
|
服务器网卡 |
Mellanox ConnectX-6 Lx |
驱动版本:MLNX_OFED_LINUX-5.4-3.2.7.2.3-rhel8.4-x86_64 固件版本: driver: mlx5_core version: 5.4-3.2.7.2.3 firmware-version: 26.31.2006 (MT_0000000531) |
Mellanox ConnectX-5 |
驱动版本:MLNX_OFED_LINUX-5.4-3.2.7.2.3-rhel8.4-x86_64 固件版本: driver: mlx5_core version: 5.4-3.2.7.2.3 firmware-version: 16.31.2006 (MT_0000000080) |
|
Mellanox ConnectX-4 Lx |
驱动版本:MLNX_OFED_LINUX-5.4-3.2.7.2.3-rhel8.4-x86_64 固件版本: driver: mlx5_core version: 5.4-3.2.7.2.3 firmware-version: 14.31.2006 (MT_2420110034) |
请客户根据实际的业务需求,制定各级设备的收敛比。建议高速端口中,2个或4个端口用于peer-link链路以满足同组Leaf下挂服务器东西向流量的转发需求,其余高速端口用作上行端口。普通速率端口连接服务器。如果收敛比不满足转发性能需求,可以采用不满挂服务器的方式。
对于S6805 & S6825 & S6850 & S9850、S9820-64H、S9820-8C设备,开启指定802.1p优先级的PFC功能后,设备会为PFC的各种门限设置一个缺省值,具体请参考1.11 参数调整简介和1.12 参数调整指导。
此缺省值在一般的组网环境下是效果较好的参数组合,不建议用户自行调整。
拥塞避免有两种配置方式,全局配置的WRED table方式和端口WRED方式,选择其中一种配置即可。两种配置方式区别如下:
· 端口WRED方式是在端口上直接配置WRED的各种参数,并开启WRED功能。端口WRED方式可以只配置RoCE队列,不同接口可以配置不同的参数,使用更为灵活。推荐此方式。
· 全局配置的WRED table方式需要在系统视图下创建WRED表,并在表中配置WRED的各种参数,然后在接口上应用WRED表。WRED表中需要为所有队列配置参数,否则未配置的队列会应用到缺省值(缺省值较小,一般情况下不是合理配置)。应用同一WRED表的接口配置一致。
请选择以下一项任务进行配置:
¡ 端口WRED方式
· 配置M-LAG
请选择以下一项任务进行配置:
¡ 端口WRED方式
· 配置VLAN接口及网关,请选择以下一项任务进行配置:
本文以Leaf 1、Leaf 2的配置为例。Leaf 3、Leaf 4的配置仅体现单挂链路的配置,其他配置与Leaf 1、Leaf 2配置类似。
Leaf 1(S6850) |
Leaf 2(S6850) |
命令说明 |
注意事项 |
priority-flow-control poolID 0 headroom 131072 |
priority-flow-control poolID 0 headroom 131072 |
设置PFC全局headroom最大可用的cell资源 |
配置为允许的最大值 |
priority-flow-control deadlock cos 3 interval 10 |
priority-flow-control deadlock cos 3 interval 10 |
设置PFC死锁检查周期,RoCE报文所在队列 |
- |
priority-flow-control deadlock precision high |
priority-flow-control deadlock precision high |
设置PFC死锁检测定时器的精度 |
配置为最高精度 |
buffer egress cell queue 3 shared ratio 100 |
buffer egress cell queue 3 shared ratio 100 |
RoCE队列出方向缓存调到最大 |
- |
buffer egress cell queue 6 shared ratio 100 |
buffer egress cell queue 6 shared ratio 100 |
CNP队列出方向缓存调到最大 |
- |
buffer apply |
buffer apply |
使Buffer配置生效 |
- |
Leaf 1(S6850) |
Leaf 2(S6850) |
命令说明 |
注意事项 |
qos wred queue table 100G-WRED-Template |
qos wred queue table 100G-WRED-Template |
配置WRED表 |
- |
queue 0 drop-level 0 low-limit 37999 high-limit 38000 discard-probability 20 queue 0 drop-level 1 low-limit 37999 high-limit 38000 discard-probability 20 queue 0 drop-level 2 low-limit 37999 high-limit 38000 discard-probability 20 |
queue 0 drop-level 0 low-limit 37999 high-limit 38000 discard-probability 20 queue 0 drop-level 1 low-limit 37999 high-limit 38000 discard-probability 20 queue 0 drop-level 2 low-limit 37999 high-limit 38000 discard-probability 20 |
配置基于队列0的WRED表的内容 |
- |
queue 1 drop-level 0 low-limit 37999 high-limit 38000 discard-probability 20 queue 1 drop-level 1 low-limit 37999 high-limit 38000 discard-probability 20 queue 1 drop-level 2 low-limit 37999 high-limit 38000 discard-probability 20 |
queue 1 drop-level 0 low-limit 37999 high-limit 38000 discard-probability 20 queue 1 drop-level 1 low-limit 37999 high-limit 38000 discard-probability 20 queue 1 drop-level 2 low-limit 37999 high-limit 38000 discard-probability 20 |
配置基于队列1的WRED表的内容 |
- |
queue 2 drop-level 0 low-limit 37999 high-limit 38000 discard-probability 20 queue 2 drop-level 1 low-limit 37999 high-limit 38000 discard-probability 20 queue 2 drop-level 2 low-limit 37999 high-limit 38000 discard-probability 20 |
queue 2 drop-level 0 low-limit 37999 high-limit 38000 discard-probability 20 queue 2 drop-level 1 low-limit 37999 high-limit 38000 discard-probability 20 queue 2 drop-level 2 low-limit 37999 high-limit 38000 discard-probability 20 |
配置基于队列2的WRED表的内容 |
- |
queue 3 drop-level 0 low-limit 1000 high-limit 2000 discard-probability 20 queue 3 drop-level 1 low-limit 1000 high-limit 2000 discard-probability 20 queue 3 drop-level 2 low-limit 1000 high-limit 2000 discard-probability 20 queue 3 weighting-constant 0 queue 3 ecn |
queue 3 drop-level 0 low-limit 1000 high-limit 2000 discard-probability 20 queue 3 drop-level 1 low-limit 1000 high-limit 2000 discard-probability 20 queue 3 drop-level 2 low-limit 1000 high-limit 2000 discard-probability 20 queue 3 weighting-constant 0 queue 3 ecn |
配置基于队列3的WRED表的内容 |
RoCE队列的low-limit和high-limit配置较小值,其他队列配置较大值 |
queue 4 drop-level 0 low-limit 37999 high-limit 38000 discard-probability 20 queue 4 drop-level 1 low-limit 37999 high-limit 38000 discard-probability 20 queue 4 drop-level 2 low-limit 37999 high-limit 38000 discard-probability 20 |
queue 4 drop-level 0 low-limit 37999 high-limit 38000 discard-probability 20 queue 4 drop-level 1 low-limit 37999 high-limit 38000 discard-probability 20 queue 4 drop-level 2 low-limit 37999 high-limit 38000 discard-probability 20 |
配置基于队列4的WRED表的内容 |
- |
queue 5 drop-level 0 low-limit 37999 high-limit 38000 discard-probability 20 queue 5 drop-level 1 low-limit 37999 high-limit 38000 discard-probability 20 queue 5 drop-level 2 low-limit 37999 high-limit 38000 discard-probability 20 |
queue 5 drop-level 0 low-limit 37999 high-limit 38000 discard-probability 20 queue 5 drop-level 1 low-limit 37999 high-limit 38000 discard-probability 20 queue 5 drop-level 2 low-limit 37999 high-limit 38000 discard-probability 20 |
配置基于队列5的WRED表的内容 |
- |
queue 6 drop-level 0 low-limit 37999 high-limit 38000 discard-probability 20 queue 6 drop-level 1 low-limit 37999 high-limit 38000 discard-probability 20 queue 6 drop-level 2 low-limit 37999 high-limit 38000 discard-probability 20 queue 6 ecn |
queue 6 drop-level 0 low-limit 37999 high-limit 38000 discard-probability 20 queue 6 drop-level 1 low-limit 37999 high-limit 38000 discard-probability 20 queue 6 drop-level 2 low-limit 37999 high-limit 38000 discard-probability 20 queue 6 ecn |
配置基于队列6的WRED表的内容 |
- |
queue 7 drop-level 0 low-limit 37999 high-limit 38000 discard-probability 20 queue 7 drop-level 1 low-limit 37999 high-limit 38000 discard-probability 20 queue 7 drop-level 2 low-limit 37999 high-limit 38000 discard-probability 20 |
queue 7 drop-level 0 low-limit 37999 high-limit 38000 discard-probability 20 queue 7 drop-level 1 low-limit 37999 high-limit 38000 discard-probability 20 queue 7 drop-level 2 low-limit 37999 high-limit 38000 discard-probability 20 |
配置基于队列7的WRED表的内容 |
- |
qos wred queue table 25G-WRED-Template |
qos wred queue table 25G-WRED-Template |
配置WRED表 |
- |
queue 0 drop-level 0 low-limit 37999 high-limit 38000 discard-probability 20 queue 0 drop-level 1 low-limit 37999 high-limit 38000 discard-probability 20 queue 0 drop-level 2 low-limit 37999 high-limit 38000 discard-probability 20 |
queue 0 drop-level 0 low-limit 37999 high-limit 38000 discard-probability 20 queue 0 drop-level 1 low-limit 37999 high-limit 38000 discard-probability 20 queue 0 drop-level 2 low-limit 37999 high-limit 38000 discard-probability 20 |
配置基于队列0的WRED表的内容 |
- |
queue 1 drop-level 0 low-limit 37999 high-limit 38000 discard-probability 20 queue 1 drop-level 1 low-limit 37999 high-limit 38000 discard-probability 20 queue 1 drop-level 2 low-limit 37999 high-limit 38000 discard-probability 20 |
queue 1 drop-level 0 low-limit 37999 high-limit 38000 discard-probability 20 queue 1 drop-level 1 low-limit 37999 high-limit 38000 discard-probability 20 queue 1 drop-level 2 low-limit 37999 high-limit 38000 discard-probability 20 |
配置基于队列1的WRED表的内容 |
- |
queue 2 drop-level 0 low-limit 37999 high-limit 38000 discard-probability 20 queue 2 drop-level 1 low-limit 37999 high-limit 38000 discard-probability 20 queue 2 drop-level 2 low-limit 37999 high-limit 38000 discard-probability 20 |
queue 2 drop-level 0 low-limit 37999 high-limit 38000 discard-probability 20 queue 2 drop-level 1 low-limit 37999 high-limit 38000 discard-probability 20 queue 2 drop-level 2 low-limit 37999 high-limit 38000 discard-probability 20 |
配置基于队列2的WRED表的内容 |
- |
queue 3 drop-level 0 low-limit 400 high-limit 1625 discard-probability 20 queue 3 drop-level 1 low-limit 400 high-limit 1625 discard-probability 20 queue 3 drop-level 2 low-limit 400 high-limit 1625 discard-probability 20 queue 3 weighting-constant 0 queue 3 ecn |
queue 3 drop-level 0 low-limit 400 high-limit 1625 discard-probability 20 queue 3 drop-level 1 low-limit 400 high-limit 1625 discard-probability 20 queue 3 drop-level 2 low-limit 400 high-limit 1625 discard-probability 20 queue 3 weighting-constant 0 queue 3 ecn |
配置基于队列3的WRED表的内容 |
RoCE队列的low-limit和high-limit配置较小值,其他队列配置较大值 |
queue 4 drop-level 0 low-limit 37999 high-limit 38000 discard-probability 20 queue 4 drop-level 1 low-limit 37999 high-limit 38000 discard-probability 20 queue 4 drop-level 2 low-limit 37999 high-limit 38000 discard-probability 20 |
queue 4 drop-level 0 low-limit 37999 high-limit 38000 discard-probability 20 queue 4 drop-level 1 low-limit 37999 high-limit 38000 discard-probability 20 queue 4 drop-level 2 low-limit 37999 high-limit 38000 discard-probability 20 |
配置基于队列4的WRED表的内容 |
- |
queue 5 drop-level 0 low-limit 37999 high-limit 38000 discard-probability 20 queue 5 drop-level 1 low-limit 37999 high-limit 38000 discard-probability 20 queue 5 drop-level 2 low-limit 37999 high-limit 38000 discard-probability 20 |
queue 5 drop-level 0 low-limit 37999 high-limit 38000 discard-probability 20 queue 5 drop-level 1 low-limit 37999 high-limit 38000 discard-probability 20 queue 5 drop-level 2 low-limit 37999 high-limit 38000 discard-probability 20 |
配置基于队列5的WRED表的内容 |
- |
queue 6 drop-level 0 low-limit 37999 high-limit 38000 discard-probability 20 queue 6 drop-level 1 low-limit 37999 high-limit 38000 discard-probability 20 queue 6 drop-level 2 low-limit 37999 high-limit 38000 discard-probability 20 queue 6 ecn |
queue 6 drop-level 0 low-limit 37999 high-limit 38000 discard-probability 20 queue 6 drop-level 1 low-limit 37999 high-limit 38000 discard-probability 20 queue 6 drop-level 2 low-limit 37999 high-limit 38000 discard-probability 20 queue 6 ecn |
配置基于队列6的WRED表的内容 |
- |
queue 7 drop-level 0 low-limit 37999 high-limit 38000 discard-probability 20 queue 7 drop-level 1 low-limit 37999 high-limit 38000 discard-probability 20 queue 7 drop-level 2 low-limit 37999 high-limit 38000 discard-probability 20 |
queue 7 drop-level 0 low-limit 37999 high-limit 38000 discard-probability 20 queue 7 drop-level 1 low-limit 37999 high-limit 38000 discard-probability 20 queue 7 drop-level 2 low-limit 37999 high-limit 38000 discard-probability 20 |
配置基于队列7的WRED表的内容 |
- |
接口配置PFC功能时,Leaf设备连接服务器的接口需要信任报文自带的802.1p优先级或DSCP优先级,Leaf和Spine互连的三层接口需要信任报文自带的DSCP优先级。
建议配置服务器发出的报文携带DSCP优先级并能根据DSCP值入队列,此时Leaf设备连接服务器的接口配置信任报文的DSCP优先级。如果服务器不支持携带DSCP优先级,Leaf设备连接服务器的接口配置信任报文自带的802.1p优先级并需配置DSCP和Dot1p映射关系(必须为所有报文涉及的DSCP配置映射关系)。
本节描述接口信任报文自带的802.1p优先级时,DSCP和Dot1p映射关系的配置方法。如果采用接口信任DSCP优先级的方式,则不需要配置本节内容。
Leaf 1(S6850) |
Leaf 2(S6850) |
命令说明 |
注意事项 |
qos map-table dot1p-lp import 0 export 0 import 1 export 1 import 2 export 2 |
qos map-table dot1p-lp import 0 export 0 import 1 export 1 import 2 export 2 |
调整dot1p-lp的队列映射 |
- |
traffic classifier dot1p0 operator and |
traffic classifier dot1p0 operator and |
创建类,各规则之间为逻辑与的关系 |
如果下行口配置qos trust dscp,则不用配置此MQC |
if-match service-dot1p 0 |
if-match service-dot1p 0 |
定义匹配外层VLAN Tag 802.1p优先级的规则 |
- |
traffic classifier dot1p1 operator and |
traffic classifier dot1p1 operator and |
创建类,各规则之间为逻辑与的关系 |
- |
if-match service-dot1p 1 |
if-match service-dot1p 1 |
定义匹配外层VLAN Tag 802.1p优先级的规则 |
- |
traffic classifier dot1p2 operator and |
traffic classifier dot1p2 operator and |
创建类,各规则之间为逻辑与的关系 |
- |
if-match service-dot1p 2 |
if-match service-dot1p 2 |
定义匹配外层VLAN Tag 802.1p优先级的规则 |
- |
traffic classifier dot1p3 operator and |
traffic classifier dot1p3 operator and |
创建类,各规则之间为逻辑与的关系 |
- |
if-match service-dot1p 3 |
if-match service-dot1p 3 |
定义匹配外层VLAN Tag 802.1p优先级的规则 |
- |
traffic classifier dot1p4 operator and |
traffic classifier dot1p4 operator and |
创建类,各规则之间为逻辑与的关系 |
- |
if-match service-dot1p 4 |
if-match service-dot1p 4 |
定义匹配外层VLAN Tag 802.1p优先级的规则 |
- |
traffic classifier dot1p5 operator and |
traffic classifier dot1p5 operator and |
创建类,各规则之间为逻辑与的关系 |
- |
if-match service-dot1p 5 |
if-match service-dot1p 5 |
定义匹配外层VLAN Tag 802.1p优先级的规则 |
- |
traffic classifier dot1p5 operator and |
traffic classifier dot1p5 operator and |
创建类,各规则之间为逻辑与的关系 |
- |
if-match service-dot1p 6 |
if-match service-dot1p 6 |
定义匹配外层VLAN Tag 802.1p优先级的规则 |
- |
traffic classifier dot1p7 operator and |
traffic classifier dot1p7 operator and |
创建类,各规则之间为逻辑与的关系 |
- |
if-match service-dot1p 7 |
if-match service-dot1p 7 |
定义匹配外层VLAN Tag 802.1p优先级的规则 |
- |
traffic classifier dscp0 operator and |
traffic classifier dscp0 operator and |
创建类,各规则之间为逻辑与的关系 |
- |
if-match dscp default |
if-match dscp default |
定义匹配外层VLAN Tag 802.1p优先级的规则 |
- |
traffic classifier dscp10 operator and |
traffic classifier dscp10 operator and |
创建类,各规则之间为逻辑与的关系 |
- |
if-match dscp af11 |
if-match dscp af11 |
定义匹配DSCP的规则 |
- |
traffic classifier dscp18 operator and |
traffic classifier dscp18 operator and |
创建类,各规则之间为逻辑与的关系 |
- |
if-match dscp af21 |
if-match dscp af21 |
定义匹配DSCP的规则 |
- |
traffic classifier dscp26 operator and |
traffic classifier dscp26 operator and |
创建类,各规则之间为逻辑与的关系 |
- |
if-match dscp af31 |
if-match dscp af31 |
定义匹配DSCP的规则 |
- |
traffic classifier dscp34 operator and |
traffic classifier dscp34 operator and |
创建类,各规则之间为逻辑与的关系 |
- |
if-match dscp af41 |
if-match dscp af41 |
定义匹配DSCP的规则 |
- |
traffic classifier dscp40 operator and |
traffic classifier dscp40 operator and |
创建类,各规则之间为逻辑与的关系 |
- |
if-match dscp cs5 |
if-match dscp cs5 |
定义匹配DSCP的规则 |
- |
traffic classifier dscp48 operator and |
traffic classifier dscp48 operator and |
创建类,各规则之间为逻辑与的关系 |
- |
if-match dscp cs6 |
if-match dscp cs6 |
定义匹配DSCP的规则 |
- |
traffic classifier dscp56 operator and |
traffic classifier dscp56 operator and |
创建类,各规则之间为逻辑与的关系 |
- |
if-match dscp cs7 |
if-match dscp cs7 |
定义匹配DSCP的规则 |
- |
traffic behavior dot1p0 |
traffic behavior dot1p0 |
创建流行为 |
- |
remark dot1p 0 |
remark dot1p 0 |
重新标记报文的802.1p优先级值 |
- |
traffic behavior dot1p1 |
traffic behavior dot1p1 |
创建流行为 |
- |
remark dot1p 1 |
remark dot1p 1 |
重新标记报文的802.1p优先级值 |
- |
traffic behavior dot1p2 |
traffic behavior dot1p2 |
创建流行为 |
- |
remark dot1p 2 |
remark dot1p 2 |
重新标记报文的802.1p优先级值 |
- |
traffic behavior dot1p3 |
traffic behavior dot1p3 |
创建流行为 |
- |
remark dot1p 3 |
remark dot1p 3 |
重新标记报文的802.1p优先级值 |
- |
traffic behavior dot1p4 |
traffic behavior dot1p4 |
创建流行为 |
- |
remark dot1p 4 |
remark dot1p 4 |
重新标记报文的802.1p优先级值 |
- |
traffic behavior dot1p5 |
traffic behavior dot1p5 |
创建流行为 |
- |
remark dot1p 5 |
remark dot1p 5 |
重新标记报文的802.1p优先级值 |
- |
traffic behavior dot1p5 |
traffic behavior dot1p5 |
创建流行为 |
- |
remark dot1p 6 |
remark dot1p 6 |
重新标记报文的802.1p优先级值 |
- |
traffic behavior dot1p7 |
traffic behavior dot1p7 |
创建流行为 |
- |
remark dot1p 7 |
remark dot1p 7 |
重新标记报文的802.1p优先级值 |
- |
traffic behavior dscp-0 |
traffic behavior dscp-0 |
创建流行为 |
- |
remark dscp default |
remark dscp default |
重新标记报文的DSCP值 |
- |
traffic behavior dscp-af11 |
traffic behavior dscp-af11 |
创建流行为 |
- |
remark dscp af11 |
remark dscp af11 |
重新标记报文的DSCP值 |
- |
traffic behavior dscp-af21 |
traffic behavior dscp-af21 |
创建流行为 |
- |
remark dscp af21 |
remark dscp af21 |
重新标记报文的DSCP值 |
- |
traffic behavior dscp-af31 |
traffic behavior dscp-af31 |
创建流行为 |
- |
remark dscp af31 |
remark dscp af31 |
重新标记报文的DSCP值 |
- |
traffic behavior dscp-af41 |
traffic behavior dscp-af41 |
创建流行为 |
- |
remark dscp af41 |
remark dscp af41 |
重新标记报文的DSCP值 |
- |
traffic behavior dscp-cs5 |
traffic behavior dscp-cs5 |
创建流行为 |
- |
remark dscp cs5 |
remark dscp cs5 |
重新标记报文的DSCP值 |
- |
traffic behavior dscp-cs6 |
traffic behavior dscp-cs6 |
创建流行为 |
- |
remark dscp cs6 |
remark dscp cs6 |
重新标记报文的DSCP值 |
- |
traffic behavior dscp-cs7 |
traffic behavior dscp-cs7 |
创建流行为 |
- |
remark dscp cs7 |
remark dscp cs7 |
重新标记报文的DSCP值 |
- |
qos policy dot1p-dscp |
qos policy dot1p-dscp |
创建通用类型的QoS策略 |
- |
classifier dot1p0 behavior dscp-0 |
classifier dot1p0 behavior dscp-0 |
为类指定流行为 |
- |
classifier dot1p3 behavior dscp-af31 |
classifier dot1p3 behavior dscp-af31 |
为类指定流行为 |
- |
classifier dot1p1 behavior dscp-af11 |
classifier dot1p1 behavior dscp-af11 |
为类指定流行为 |
- |
classifier dot1p2 behavior dscp-af21 |
classifier dot1p2 behavior dscp-af21 |
为类指定流行为 |
- |
classifier dot1p4 behavior dscp-af41 |
classifier dot1p4 behavior dscp-af41 |
为类指定流行为 |
- |
classifier dot1p5 behavior dscp-cs5 |
classifier dot1p5 behavior dscp-cs5 |
为类指定流行为 |
- |
classifier dot1p5 behavior dscp-cs6 |
classifier dot1p5 behavior dscp-cs6 |
为类指定流行为 |
- |
classifier dot1p7 behavior dscp-cs7 |
classifier dot1p7 behavior dscp-cs7 |
为类指定流行为 |
- |
qos policy dscptodot1p |
qos policy dscptodot1p |
创建通用类型的QoS策略 |
- |
classifier dscp0 behavior dot1p0 |
classifier dscp0 behavior dot1p0 |
为类指定流行为 |
- |
classifier dscp26 behavior dot1p3 |
classifier dscp26 behavior dot1p3 |
为类指定流行为 |
- |
classifier dscp10 behavior dot1p1 |
classifier dscp10 behavior dot1p1 |
为类指定流行为 |
- |
classifier dscp18 behavior dot1p2 |
classifier dscp18 behavior dot1p2 |
为类指定流行为 |
- |
classifier dscp34 behavior dot1p4 |
classifier dscp34 behavior dot1p4 |
为类指定流行为 |
- |
classifier dscp40 behavior dot1p5 |
classifier dscp40 behavior dot1p5 |
为类指定流行为 |
- |
classifier dscp48 behavior dot1p5 |
classifier dscp48 behavior dot1p5 |
为类指定流行为 |
- |
classifier dscp56 behavior dot1p7 |
classifier dscp56 behavior dot1p7 |
为类指定流行为 |
- |
Leaf 1(S6850) |
Leaf 2(S6850) |
命令说明 |
注意事项 |
monitor-link group 1 |
monitor-link group 1 |
配置monitor-group |
全局配置 |
interface HundredGigE1/0/25 |
interface HundredGigE1/0/25 |
配置与Spine 1互联接口 |
- |
port link-mode route |
port link-mode route |
将以太网接口切换到三层工作模式 |
- |
ip address 172.16.2.154 255.255.255.252 |
ip address 172.16.2.158 255.255.255.252 |
配置IP |
- |
port monitor-link group 1 uplink |
port monitor-link group 1 uplink |
配置端口为monitor-link上行口 |
- |
interface HundredGigE1/0/26 |
interface HundredGigE1/0/26 |
配置与Spine 1互联接口 |
- |
port link-mode route |
port link-mode route |
将以太网接口切换到三层工作模式 |
- |
ip address 172.16.3.154 255.255.255.252 |
ip address 172.16.3.158 255.255.255.252 |
配置IP |
- |
port monitor-link group 1 uplink |
port monitor-link group 1 uplink |
配置端口为monitor-link上行口 |
- |
Leaf 1(S6850) |
Leaf 2(S6850) |
命令说明 |
注意事项 |
interface range HundredGigE1/0/25 HundredGigE1/0/26 |
interface range HundredGigE1/0/25 HundredGigE1/0/26 |
配置与Spine 1互联接口 |
- |
priority-flow-control deadlock enable |
priority-flow-control deadlock enable |
开启端口PFC死锁检测功能 |
- |
priority-flow-control enable |
priority-flow-control enable |
开启端口PFC功能 |
- |
priority-flow-control no-drop dot1p 3 |
priority-flow-control no-drop dot1p 3 |
开启RoCE报文所在队列的PFC功能 |
- |
priority-flow-control dot1p 3 headroom 491 |
priority-flow-control dot1p 3 headroom 491 |
配置3队列headroom缓存门限为491 |
|
priority-flow-control dot1p 3 reserved-buffer 17 |
priority-flow-control dot1p 3 reserved-buffer 17 |
配置PFC预留门限 |
|
priority-flow-control dot1p 3 ingress-buffer dynamic 5 |
priority-flow-control dot1p 3 ingress-buffer dynamic 5 |
配置动态反压帧触发门限 |
|
priority-flow-control dot1p 3 ingress-threshold-offset 12 |
priority-flow-control dot1p 3 ingress-threshold-offset 12 |
配置反压帧停止门限与触发门限间的偏移量 |
|
qos trust dscp |
qos trust dscp |
端口QoS信任DSCP |
- |
qos wfq byte-count |
qos wfq byte-count |
设置WFQ队列调度 |
- |
qos wfq be group 1 byte-count 15 |
qos wfq be group 1 byte-count 15 |
设置0队列调度权重 |
根据实际业务情况配置调度权重 |
qos wfq af1 group 1 byte-count 2 |
qos wfq af1 group 1 byte-count 2 |
设置1队列调度权重 |
根据实际业务情况配置调度权重 |
qos wfq af2 group 1 byte-count 2 |
qos wfq af2 group 1 byte-count 2 |
设置2队列调度权重 |
根据实际业务情况配置调度权重 |
qos wfq af3 group 1 byte-count 60 |
qos wfq af3 group 1 byte-count 60 |
设置3队列调度权重 |
根据实际业务情况配置调度权重 本文以配置无损队列和其他队列权重占比为4:1为例 |
qos wfq cs6 group sp |
qos wfq cs6 group sp |
设置6队列调度优先级为sp |
根据实际业务情况配置调度权重 |
qos wfq cs7 group sp |
qos wfq cs7 group sp |
设置7队列调度优先级为sp |
根据实际业务情况配置调度权重 |
qos wred apply 100G-WRED-Template |
qos wred apply 100G-WRED-Template |
端口使能100G的WRED表 |
- |
qos gts queue 6 cir 50000000 cbs 16000000 |
qos gts queue 6 cir 50000000 cbs 16000000 |
CNP队列限速50G |
- |
Leaf 1(S6850) |
Leaf 2(S6850) |
命令说明 |
注意事项 |
interface range HundredGigE1/0/25 HundredGigE1/0/26 |
interface range HundredGigE1/0/25 HundredGigE1/0/26 |
配置与Spine 1互联接口 |
- |
priority-flow-control deadlock enable |
priority-flow-control deadlock enable |
开启端口PFC死锁检测功能 |
- |
priority-flow-control enable |
priority-flow-control enable |
开启端口PFC功能 |
- |
priority-flow-control no-drop dot1p 3 |
priority-flow-control no-drop dot1p 3 |
开启RoCE报文所在队列的PFC功能 |
- |
priority-flow-control dot1p 3 headroom 491 |
priority-flow-control dot1p 3 headroom 491 |
配置3队列headroom缓存门限为491 |
|
priority-flow-control dot1p 3 reserved-buffer 17 |
priority-flow-control dot1p 3 reserved-buffer 17 |
配置PFC预留门限 |
|
priority-flow-control dot1p 3 ingress-buffer dynamic 5 |
priority-flow-control dot1p 3 ingress-buffer dynamic 5 |
配置动态反压帧触发门限 |
|
priority-flow-control dot1p 3 ingress-threshold-offset 12 |
priority-flow-control dot1p 3 ingress-threshold-offset 12 |
配置反压帧停止门限与触发门限间的偏移量 |
|
qos trust dscp |
qos trust dscp |
端口QoS信任DSCP |
- |
qos wfq byte-count |
qos wfq byte-count |
设置WFQ队列调度 |
- |
qos wfq be group 1 byte-count 15 |
qos wfq be group 1 byte-count 15 |
设置0队列调度权重 |
根据实际业务情况配置调度权重 |
qos wfq af1 group 1 byte-count 2 |
qos wfq af1 group 1 byte-count 2 |
设置1队列调度权重 |
根据实际业务情况配置调度权重 |
qos wfq af2 group 1 byte-count 2 |
qos wfq af2 group 1 byte-count 2 |
设置2队列调度权重 |
根据实际业务情况配置调度权重 |
qos wfq af3 group 1 byte-count 60 |
qos wfq af3 group 1 byte-count 60 |
设置3队列调度权重 |
根据实际业务情况配置调度权重 本文以配置无损队列和其他队列权重占比为4:1为例 |
qos wfq cs6 group sp |
qos wfq cs6 group sp |
设置6队列调度优先级为sp |
根据实际业务情况配置调度权重 |
qos wfq cs7 group sp |
qos wfq cs7 group sp |
设置7队列调度优先级为sp |
根据实际业务情况配置调度权重 |
qos wred queue 3 drop-level 0 low-limit 1000 high-limit 2000 discard-probability 20 qos wred queue 3 drop-level 1 low-limit 1300 high-limit 2100 discard-probability 20 qos wred queue 3 drop-level 2 low-limit 1300 high-limit 2100 discard-probability 20 |
qos wred queue 3 drop-level 0 low-limit 1000 high-limit 2000 discard-probability 20 qos wred queue 3 drop-level 1 low-limit 1300 high-limit 2100 discard-probability 20 qos wred queue 3 drop-level 2 low-limit 1300 high-limit 2100 discard-probability 20 |
设置RoCE队列的ECN水线值 |
RoCE队列的low-limit和high-limit配置较小值,其他队列配置较大值 |
qos wred queue 3 weighting-constant 0 |
qos wred queue 3 weighting-constant 0 |
设置RoCE队列平均队列长度指数 |
- |
qos wred queue 3 ecn |
qos wred queue 3 ecn |
开启RoCE队列ECN |
- |
qos wred queue 6 drop-level 0 low-limit 37999 high-limit 38000 discard-probability 20 qos wred queue 6 drop-level 1 low-limit 37999 high-limit 38000 discard-probability 20 qos wred queue 6 drop-level 2 low-limit 37999 high-limit 38000 discard-probability 20 |
qos wred queue 6 drop-level 0 low-limit 37999 high-limit 38000 discard-probability 20 qos wred queue 6 drop-level 1 low-limit 37999 high-limit 38000 discard-probability 20 qos wred queue 6 drop-level 2 low-limit 37999 high-limit 38000 discard-probability 20 |
设置CNP队列的ECN水线值 |
- |
qos wred queue 6 ecn |
qos wred queue 6 ecn |
开启CNP队列ECN |
- |
qos gts queue 6 cir 50000000 cbs 16000000 |
qos gts queue 6 cir 50000000 cbs 16000000 |
CNP队列限速50G |
- |
Leaf 1(S6850) |
Leaf 2(S6850) |
命令说明 |
注意事项 |
interface Bridge-Aggregation1000 |
interface Bridge-Aggregation1000 |
创建聚合1000作为peer-link接口 |
- |
link-aggregation mode dynamic |
link-aggregation mode dynamic |
配置聚合组工作在动态聚合模式下,同时开启LACP协议 |
- |
quit |
quit |
- |
- |
interface range HundredGigE1/0/29 to HundredGigE1/0/32 |
interface range HundredGigE1/0/29 to HundredGigE1/0/32 |
进入接口批量配置视图 |
- |
port link-aggregation group 1000 |
port link-aggregation group 1000 |
将接口加入二层聚合组 1000 |
- |
quit |
quit |
- |
- |
interface Bridge-Aggregation1000 |
interface Bridge-Aggregation1000 |
- |
- |
port m-lag peer-link 1 |
port m-lag peer-link 1 |
配置该接口为peer-link接口 |
- |
quit |
quit |
- |
- |
ip vpn-instance M-LAGKeepalive |
ip vpn-instance M-LAGKeepalive |
配置VPN用于M-LAG Keepalive端口 |
- |
quit |
quit |
quit |
- |
interface Route-Aggregation1000 |
interface Route-Aggregation1000 |
创建三层聚合口1000作为Keepalive端口 |
- |
ip binding vpn-instance M-LAGKeepalive |
ip binding vpn-instance M-LAGKeepalive |
端口绑定VPN实例 |
- |
ip address 1.1.1.1 255.255.255.252 |
ip address 1.1.1.2 255.255.255.252 |
配置Keepalive接口地址 |
- |
link-aggregation mode dynamic |
link-aggregation mode dynamic |
配置动态聚合模式 |
- |
quit |
quit |
- |
- |
interface Twenty-FiveGigE1/0/55 |
interface Twenty-FiveGigE1/0/55 |
配置该接口作为Keepalive接口的成员口 |
- |
port link-mode route |
port link-mode route |
配置端口为路由口 |
- |
port link-aggregation group 1000 |
port link-aggregation group 1000 |
将端口加入三层聚合组1000 |
- |
quit |
quit |
- |
- |
interface Twenty-FiveGigE1/0/56 |
interface Twenty-FiveGigE1/0/56 |
配置该接口作为Keepalive接口的成员口 |
- |
port link-mode route |
port link-mode route |
配置端口为路由口 |
- |
port link-aggregation group 1000 |
port link-aggregation group 1000 |
将端口加入三层聚合组1000 |
- |
quit |
quit |
- |
- |
m-lag role priority 50 |
m-lag role priority 100 |
指定M-LAG设备的角色优先级 |
全局配置 |
m-lag system-mac 2001-0000-0018 |
m-lag system-mac 2001-0000-0018 |
配置M-LAG系统MAC地址 |
同一M-LAG组中,所有M-LAG设备的系统MAC地址必须相同 |
m-lag system-number 1 |
m-lag system-number 2 |
配置M-LAG系统编号 |
M-LAG系统中不同M-LAG设备的系统编号必须不同 |
m-lag system-priority 110 |
m-lag system-priority 110 |
(可选)配置M-LAG系统优先级 |
为了组成M-LAG组,同一M-LAG组中,所有M-LAG设备的系统优先级必须相同 缺省情况下,M-LAG系统优先级为32768。该数值越小,优先级越高。 |
m-lag keepalive ip destination 1.1.1.2 source 1.1.1.1 vpn-instance M-LAGKeepalive |
m-lag keepalive ip destination 1.1.1.1 source 1.1.1.2 vpn-instance M-LAGKeepalive |
配置Keepalive报文的目的IP地址和源IP地址 |
若IP地址所在的接口为非M-LAG保留接口,则配置将接口配置为保留接口 |
m-lag mad exclude interface LoopBack1 |
m-lag mad exclude interface LoopBack1 |
配置接口在M-LAG系统分裂后不处于M-LAG MAD DOWN状态 |
在Loopback1端口配置 |
m-lag mad exclude interface Route-Aggregation1000 |
m-lag mad exclude interface Route-Aggregation1000 |
配置接口在M-LAG系统分裂后不处于M-LAG MAD DOWN状态 |
- |
m-lag mad exclude interface Twenty-FiveGigE1/0/55 |
m-lag mad exclude interface Twenty-FiveGigE1/0/55 |
配置接口在M-LAG系统分裂后不处于M-LAG MAD DOWN状态 |
在keepalive端口配置 |
m-lag mad exclude interface Twenty-FiveGigE1/0/56 |
m-lag mad exclude interface Twenty-FiveGigE1/0/56 |
配置接口在M-LAG系统分裂后不处于M-LAG MAD DOWN状态 |
在keepalive端口配置 |
m-lag mad exclude interface M-GigabitEthernet0/0/0 |
m-lag mad exclude interface M-GigabitEthernet0/0/0 |
配置接口在M-LAG系统分裂后不处于M-LAG MAD DOWN状态 |
在管理口配置 |
Leaf 1(S6850) |
Leaf 2(S6850) |
命令说明 |
注意事项 |
interface range HundredGigE1/0/29 to HundredGigE1/0/32 |
interface range HundredGigE1/0/29 to HundredGigE1/0/32 |
进入peer-link链路物理接口批量配置视图 |
- |
qos trust dscp |
qos trust dscp |
端口QoS信任DSCP |
- |
qos wfq byte-count |
qos wfq byte-count |
设置WFQ队列调度 |
- |
qos wfq be group 1 byte-count 15 |
qos wfq be group 1 byte-count 15 |
设置0队列调度权重 |
根据实际业务情况配置调度权重 |
qos wfq af1 group 1 byte-count 2 |
qos wfq af1 group 1 byte-count 2 |
设置1队列调度权重 |
根据实际业务情况配置调度权重 |
qos wfq af2 group 1 byte-count 2 |
qos wfq af2 group 1 byte-count 2 |
设置2队列调度权重 |
根据实际业务情况配置调度权重 |
qos wfq af3 group 1 byte-count 60 |
qos wfq af3 group 1 byte-count 60 |
设置3队列调度权重 |
根据实际业务情况配置调度权重 本文以配置无损队列和其他队列权重占比为4:1为例 |
qos wfq cs6 group sp |
qos wfq cs6 group sp |
设置6队列调度优先级为sp |
根据实际业务情况配置调度权重 |
qos wfq cs7 group sp |
qos wfq cs7 group sp |
设置7队列调度优先级为sp |
根据实际业务情况配置调度权重 |
qos wred apply 100G-WRED-Template |
qos wred apply 100G-WRED-Template |
端口使能100G的WRED表 |
- |
qos gts queue 6 cir 50000000 cbs 16000000 |
qos gts queue 6 cir 50000000 cbs 16000000 |
CNP队列限速50G |
- |
quit |
quit |
- |
- |
Leaf 1(S6850) |
Leaf 2(S6850) |
命令说明 |
注意事项 |
interface range HundredGigE1/0/29 to HundredGigE1/0/32 |
interface range HundredGigE1/0/29 to HundredGigE1/0/32 |
进入peer-link链路物理接口批量配置视图 |
- |
qos trust dscp |
qos trust dscp |
端口QoS信任DSCP |
- |
qos wfq byte-count |
qos wfq byte-count |
设置WFQ队列调度 |
- |
qos wfq be group 1 byte-count 15 |
qos wfq be group 1 byte-count 15 |
设置0队列调度权重 |
根据实际业务情况配置调度权重 |
qos wfq af1 group 1 byte-count 2 |
qos wfq af1 group 1 byte-count 2 |
设置1队列调度权重 |
根据实际业务情况配置调度权重 |
qos wfq af2 group 1 byte-count 2 |
qos wfq af2 group 1 byte-count 2 |
设置2队列调度权重 |
根据实际业务情况配置调度权重 |
qos wfq af3 group 1 byte-count 60 |
qos wfq af3 group 1 byte-count 60 |
设置3队列调度权重 |
根据实际业务情况配置调度权重 本文以配置无损队列和其他队列权重占比为4:1为例 |
qos wfq cs6 group sp |
qos wfq cs6 group sp |
设置6队列调度优先级为sp |
根据实际业务情况配置调度权重 |
qos wfq cs7 group sp |
qos wfq cs7 group sp |
设置7队列调度优先级为sp |
根据实际业务情况配置调度权重 |
qos wred queue 3 drop-level 0 low-limit 1000 high-limit 2000 discard-probability 20 qos wred queue 3 drop-level 1 low-limit 1300 high-limit 2100 discard-probability 20 qos wred queue 3 drop-level 2 low-limit 1300 high-limit 2100 discard-probability 20 |
qos wred queue 3 drop-level 0 low-limit 1000 high-limit 2000 discard-probability 20 qos wred queue 3 drop-level 1 low-limit 1300 high-limit 2100 discard-probability 20 qos wred queue 3 drop-level 2 low-limit 1300 high-limit 2100 discard-probability 20 |
设置RoCE队列的ECN水线值 |
RoCE队列的low-limit和high-limit配置较小值,其他队列配置较大值 |
qos wred queue 3 weighting-constant 0 |
qos wred queue 3 weighting-constant 0 |
设置RoCE队列平均队列长度指数 |
- |
qos wred queue 3 ecn |
qos wred queue 3 ecn |
开启RoCE队列ECN |
- |
qos wred queue 6 drop-level 0 low-limit 37999 high-limit 38000 discard-probability 20 qos wred queue 6 drop-level 1 low-limit 37999 high-limit 38000 discard-probability 20 qos wred queue 6 drop-level 2 low-limit 37999 high-limit 38000 discard-probability 20 |
qos wred queue 6 drop-level 0 low-limit 37999 high-limit 38000 discard-probability 20 qos wred queue 6 drop-level 1 low-limit 37999 high-limit 38000 discard-probability 20 qos wred queue 6 drop-level 2 low-limit 37999 high-limit 38000 discard-probability 20 |
设置CNP队列的ECN水线值 |
- |
qos wred queue 6 ecn |
qos wred queue 6 ecn |
开启CNP队列ECN |
- |
qos gts queue 6 cir 50000000 cbs 16000000 |
qos gts queue 6 cir 50000000 cbs 16000000 |
CNP队列限速50G |
- |
quit |
quit |
- |
- |
本节以Leaf 1、Leaf 2 Twenty-FiveGigE1/0/1端口的配置为例,其他端口配置类似,具体配置略。
Leaf 1(S6850) |
Leaf 2(S6850) |
命令说明 |
注意事项 |
interface bridge-aggregation 100 |
interface bridge-aggregation 100 |
创建聚合组100 |
- |
link-aggregation mode dynamic |
link-aggregation mode dynamic |
设置聚合模式为动态聚合 |
- |
port m-lag group 1 |
port m-lag group 1 |
配置接口为M-LAG接口1 |
- |
quit |
quit |
- |
- |
interface Twenty-FiveGigE1/0/1 |
interface Twenty-FiveGigE1/0/1 |
配置与服务器互联端口 |
- |
port link-aggregation group 100 |
port link-aggregation group 100 |
将物理口加入二层聚合组100中 |
- |
broadcast-suppression 1 |
broadcast-suppression 1 |
设置端口广播抑制 |
- |
multicast-suppression 1 |
multicast-suppression 1 |
设置端口组播抑制 |
- |
stp edged-port |
stp edged-port |
配置端口为边缘端口 |
- |
port monitor-link group 1 downlink |
port monitor-link group 1 downlink |
配置端口为monitor-link的下行口 |
- |
quit |
quit |
- |
- |
interface bridge-aggregation 100 |
interface bridge-aggregation 100 |
- |
- |
port link-type trunk |
port link-type trunk |
配置端口的链路类型为Trunk类型 |
- |
undo port trunk permit vlan 1 |
undo port trunk permit vlan 1 |
配置Trunk端口不允许VLAN 1的报文通过 |
- |
port trunk permit vlan 1100 to 1500 |
port trunk permit vlan 1100 to 1500 |
配置Trunk端口允许通过的VLAN为VLAN 1100~1500 |
- |
quit |
quit |
- |
- |
Leaf 3(S6850) |
Leaf 3(S6850) |
命令说明 |
注意事项 |
interface range Twenty-FiveGigE1/0/1 Twenty-FiveGigE1/0/2 |
interface range Twenty-FiveGigE1/0/1 Twenty-FiveGigE1/0/2 |
配置与服务器互联端口 |
- |
port link-type trunk |
port link-type trunk |
配置端口的链路类型为Trunk类型 |
- |
undo port trunk permit vlan 1 |
undo port trunk permit vlan 1 |
配置Trunk端口不允许VLAN 1的报文通过 |
- |
port trunk permit vlan 1100 to 1500 |
port trunk permit vlan 1100 to 1500 |
配置Trunk端口允许通过的VLAN为VLAN 1100~1500 |
- |
broadcast-suppression 1 |
broadcast-suppression 1 |
设置端口广播抑制 |
- |
multicast-suppression 1 |
multicast-suppression 1 |
设置端口组播抑制 |
- |
stp edged-port |
stp edged-port |
配置端口为边缘端口 |
- |
port monitor-link group 1 downlink |
port monitor-link group 1 downlink |
配置端口为monitor-link的下行口 |
- |
quit |
quit |
- |
- |
interface bridge-aggregation 100 |
interface bridge-aggregation 100 |
创建聚合组100 |
创建一个空M-LAG接口用于规避单挂链路的使用限制 单挂链路的使用限制请参考“M-LAG组网推荐方案”中的“单挂(即服务器通过非M-LAG接口接入)接入使用限制” |
link-aggregation mode dynamic |
link-aggregation mode dynamic |
设置聚合模式为动态聚合 |
- |
port m-lag group 1 |
port m-lag group 1 |
配置接口为M-LAG接口1 |
- |
port link-type trunk |
port link-type trunk |
配置端口的链路类型为Trunk类型 |
- |
undo port trunk permit vlan 1 |
undo port trunk permit vlan 1 |
配置Trunk端口不允许VLAN 1的报文通过 |
- |
port trunk permit vlan 1100 to 1500 |
port trunk permit vlan 1100 to 1500 |
配置Trunk端口允许通过的VLAN为VLAN 1100~1500 |
- |
quit |
quit |
- |
- |
Leaf 1(S6850) |
Leaf 2(S6850) |
命令说明 |
注意事项 |
interface Twenty-FiveGigE1/0/1 |
interface Twenty-FiveGigE1/0/1 |
配置与服务器互联端口 |
- |
priority-flow-control deadlock enable |
priority-flow-control deadlock enable |
开启端口PFC死锁检测功能 |
- |
priority-flow-control enable |
priority-flow-control enable |
开启端口PFC功能 |
|
priority-flow-control no-drop dot1p 3 |
priority-flow-control no-drop dot1p 3 |
开启RoCE报文所在队列的PFC功能 |
- |
priority-flow-control dot1p 3 headroom 125 |
priority-flow-control dot1p 3 headroom 125 |
配置3队列headroom缓存门限为125 |
|
priority-flow-control dot1p 3 reserved-buffer 17 |
priority-flow-control dot1p 3 reserved-buffer 17 |
配置PFC预留门限 |
|
priority-flow-control dot1p 3 ingress-buffer dynamic 5 |
priority-flow-control dot1p 3 ingress-buffer dynamic 5 |
配置动态反压帧触发门限 |
|
priority-flow-control dot1p 3 ingress-threshold-offset 12 |
priority-flow-control dot1p 3 ingress-threshold-offset 12 |
配置反压帧停止门限与触发门限间的偏移量 |
|
qos trust dot1p |
qos trust dot1p |
端口QoS信任Dot1p |
- |
qos wfq byte-count |
qos wfq byte-count |
设置WFQ队列调度 |
- |
qos wfq be group 1 byte-count 15 |
qos wfq be group 1 byte-count 15 |
设置0队列调度权重 |
根据实际业务情况配置调度权重 |
qos wfq af1 group 1 byte-count 2 |
qos wfq af1 group 1 byte-count 2 |
设置1队列调度权重 |
根据实际业务情况配置调度权重 |
qos wfq af2 group 1 byte-count 2 |
qos wfq af2 group 1 byte-count 2 |
设置2队列调度权重 |
根据实际业务情况配置调度权重 |
qos wfq af3 group 1 byte-count 60 |
qos wfq af3 group 1 byte-count 60 |
设置3队列调度权重 |
根据实际业务情况配置调度权重 本文以配置无损队列和其他队列权重占比为4:1为例 |
qos wfq cs6 group sp |
qos wfq cs6 group sp |
设置6队列调度优先级为sp |
根据实际业务情况配置调度权重 |
qos wfq cs7 group sp |
qos wfq cs7 group sp |
设置7队列调度优先级为sp |
根据实际业务情况配置调度权重 |
qos apply policy dot1p-dscp inbound |
qos apply policy dot1p-dscp inbound |
入方向配置mqc进行ddot1p-dscp的标记 |
- |
qos apply policy dscptodot1p outbound |
qos apply policy dscptodot1p outbound |
出方向配置mqc进行dscp-dot1p的标记 |
- |
qos wred apply 25G-WRED-Template |
qos wred apply 25G-WRED-Template |
端口使能25G的WRED表 |
- |
qos gts queue 6 cir 12500000 cbs 16000000 |
qos gts queue 6 cir 12500000 cbs 16000000 |
CNP队列限速12.5G |
- |
Leaf 1(S6850) |
Leaf 2(S6850) |
命令说明 |
注意事项 |
interface Twenty-FiveGigE1/0/1 |
interface Twenty-FiveGigE1/0/1 |
配置与服务器互联端口 |
- |
priority-flow-control deadlock enable |
priority-flow-control deadlock enable |
开启端口PFC死锁检测功能 |
- |
priority-flow-control enable |
priority-flow-control enable |
开启端口PFC功能 |
- |
priority-flow-control no-drop dot1p 3 |
priority-flow-control no-drop dot1p 3 |
开启RoCE报文所在队列的PFC功能 |
- |
priority-flow-control dot1p 3 headroom 125 |
priority-flow-control dot1p 3 headroom 125 |
配置3队列headroom缓存门限为125 |
|
priority-flow-control dot1p 3 reserved-buffer 17 |
priority-flow-control dot1p 3 reserved-buffer 17 |
配置PFC预留门限 |
|
priority-flow-control dot1p 3 ingress-buffer dynamic 5 |
priority-flow-control dot1p 3 ingress-buffer dynamic 5 |
配置动态反压帧触发门限 |
|
priority-flow-control dot1p 3 ingress-threshold-offset 12 |
priority-flow-control dot1p 3 ingress-threshold-offset 12 |
配置反压帧停止门限与触发门限间的偏移量 |
|
qos trust dot1p |
qos trust dot1p |
端口QoS信任Dot1p |
- |
qos wfq byte-count |
qos wfq byte-count |
设置WFQ队列调度 |
- |
qos wfq be group 1 byte-count 15 |
qos wfq be group 1 byte-count 15 |
设置0队列调度权重 |
根据实际业务情况配置调度权重 |
qos wfq af1 group 1 byte-count 2 |
qos wfq af1 group 1 byte-count 2 |
设置1队列调度权重 |
根据实际业务情况配置调度权重 |
qos wfq af2 group 1 byte-count 2 |
qos wfq af2 group 1 byte-count 2 |
设置2队列调度权重 |
根据实际业务情况配置调度权重 |
qos wfq af3 group 1 byte-count 60 |
qos wfq af3 group 1 byte-count 60 |
设置3队列调度权重 |
根据实际业务情况配置调度权重 本文以配置无损队列和其他队列权重占比为4:1为例 |
qos wfq cs6 group sp |
qos wfq cs6 group sp |
设置6队列调度优先级为sp |
根据实际业务情况配置调度权重 |
qos wfq cs7 group sp |
qos wfq cs7 group sp |
设置7队列调度优先级为sp |
根据实际业务情况配置调度权重 |
qos apply policy dot1p-dscp inbound |
qos apply policy dot1p-dscp inbound |
入方向配置mqc进行ddot1p-dscp的标记 |
- |
qos apply policy dscptodot1p outbound |
qos apply policy dscptodot1p outbound |
出方向配置mqc进行dscp-dot1p的标记 |
- |
qos wred queue 3 drop-level 0 low-limit 400 high-limit 1625 discard-probability 20 qos wred queue 3 drop-level 1 low-limit 400 high-limit 1625 discard-probability 20 qos wred queue 3 drop-level 2 low-limit 400 high-limit 1625 discard-probability 20 |
qos wred queue 3 drop-level 0 low-limit 400 high-limit 1625 discard-probability 20 qos wred queue 3 drop-level 1 low-limit 400 high-limit 1625 discard-probability 20 qos wred queue 3 drop-level 2 low-limit 400 high-limit 1625 discard-probability 20 |
设置RoCE队列的水线值 |
RoCE队列的low-limit和high-limit配置较小值,其他队列配置较大值 |
qos wred queue 3 ecn |
qos wred queue 3 ecn |
开启RoCE队列的ECN |
- |
qos wred queue 3 weighting-constant 0 |
qos wred queue 3 weighting-constant 0 |
设置RoCE队列的平均队列长度指数 |
- |
qos wred queue 6 drop-level 0 low-limit 37999 high-limit 38000 discard-probability 20 qos wred queue 6 drop-level 1 low-limit 37999 high-limit 38000 discard-probability 20 qos wred queue 6 drop-level 2 low-limit 37999 high-limit 38000 discard-probability 20 |
qos wred queue 6 drop-level 0 low-limit 37999 high-limit 38000 discard-probability 20 qos wred queue 6 drop-level 1 low-limit 37999 high-limit 38000 discard-probability 20 qos wred queue 6 drop-level 2 low-limit 37999 high-limit 38000 discard-probability 20 |
设置CNP队列水线值 |
-- |
qos wred queue 6 ecn |
qos wred queue 6 ecn |
开启CNP队列ECN |
- |
qos gts queue 6 cir 12500000 cbs 16000000 |
qos gts queue 6 cir 12500000 cbs 16000000 |
CNP队列限速12.5G |
- |
Leaf 1(S6850) |
Leaf 2(S6850) |
命令说明 |
注意事项 |
bgp 64636 |
bgp 64636 |
启动BGP实例 |
- |
router-id 50.50.255.41 |
router-id 50.50.255.42 |
为指定的BGP实例配置BGP路由器的Router ID |
如果要在BGP实例下运行BGP协议,则必须为BGP实例指定Router ID。如果没有指定Router ID,则缺省使用全局Router ID |
group Spine external |
group Spine external |
创建EBGP对等体组 |
- |
peer Spine as-number 64637 |
peer Spine as-number 64637 |
指定对等体的AS号 |
- |
peer 172.16.2.153 group Spine |
peer 172.16.2.157 group Spine |
向对等体组中添加指定的对等体,即Spine |
- |
peer 172.16.3.153 group Spine |
peer 172.16.3.157 group Spine |
向对等体组中添加指定的对等体,即Spine |
- |
address-family ipv4 unicast |
address-family ipv4 unicast |
进入IPv4地址族 |
- |
balance 32 |
balance 32 |
配置BGP负载分担的路由条数 |
|
network 55.50.138.0 255.255.255.128 |
network 55.50.138.0 255.255.255.128 |
BGP发布指定网段路由 |
|
network 55.50.153.128 255.255.255.128 |
network 55.50.153.128 255.255.255.128 |
BGP发布指定网段路由 |
- |
network 55.50.250.0 255.255.255.192 |
network 55.50.250.0 255.255.255.192 |
BGP发布指定网段路由 |
- |
network 55.50.255.41 255.255.255.255 |
network 55.50.255.41 255.255.255.255 |
BGP发布指定网段路由 |
- |
peer Spine enable |
peer Spine enable |
允许本地路由器与指定对等体交换IPv4单播路由信息 |
- |
Leaf 1(S6850) |
Leaf 2(S6850) |
命令说明 |
注意事项 |
interface Vlan-interface1121 |
interface Vlan-interface1121 |
配置VLAN接口 |
- |
ip address 55.50.138.124 255.255.255.128 |
ip address 55.50.138.124 255.255.255.128 |
配置相同的IP地址 |
- |
mac-address 0001-0001-0001 |
mac-address 0001-0001-0001 |
配置相同的MAC地址 |
|
dhcp select relay |
dhcp select relay |
配置接口工作在DHCP中继模式 |
- |
dhcp relay server-address 55.50.128.12 |
dhcp relay server-address 55.50.128.12 |
指定DHCP服务器的地址 |
- |
quit |
quit |
- |
- |
m-lag mad exclude interface Vlan-interface 1121 |
m-lag mad exclude interface Vlan-interface 1121 |
配置接口在M-LAG系统分裂后不处于M-LAG MAD DOWN状态 |
在下行口VLAN接口配置 |
Leaf 1(S6850) |
Leaf 2(S6850) |
命令说明 |
注意事项 |
interface Vlan-interface1121 |
interface Vlan-interface1121 |
配置VLAN接口 |
- |
ip address 55.50.138.124 255.255.255.128 |
ip address 55.50.138.123 255.255.255.128 |
配置IP地址 |
- |
vrrp vrid 1 virtual-ip 55.50.138.125 |
vrrp vrid 1 virtual-ip 55.50.138.125 |
配置VRRP虚IP |
|
vrrp vrid 1 priority 150 |
vrrp vrid 1 priority 120 |
配置VRRP优先级 |
|
dhcp select relay |
dhcp select relay |
配置接口工作在DHCP中继模式 |
- |
dhcp relay server-address 55.50.128.12 |
dhcp relay server-address 55.50.128.12 |
指定DHCP服务器的地址 |
- |
quit |
quit |
- |
- |
m-lag mad exclude interface Vlan-interface 1121 |
m-lag mad exclude interface Vlan-interface 1121 |
配置接口在M-LAG系统分裂后不处于M-LAG MAD DOWN状态 |
在下行口VLAN接口配置 |
请选择以下一项任务进行配置:
¡ 端口WRED方式
Spine 1 |
Spine 2 |
命令说明 |
注意事项 |
priority-flow-control poolID 0 headroom 130000 |
priority-flow-control poolID 0 headroom 130000 |
设置PFC全局headroom最大可用的cell资源 |
配置为允许的最大值 |
priority-flow-control deadlock cos 3 interval 10 |
priority-flow-control deadlock cos 3 interval 10 |
设置PFC死锁检查周期,RoCE报文所在队列 |
- |
priority-flow-control deadlock precision high |
priority-flow-control deadlock precision high |
设置PFC死锁检测定时器的精度 |
配置为最高精度 |
buffer egress cell queue 3 shared ratio 100 |
buffer egress cell queue 3 shared ratio 100 |
RoCE队列出方向缓存调到最大 |
- |
buffer egress cell queue 6 shared ratio 100 |
buffer egress cell queue 6 shared ratio 100 |
CNP队列出方向缓存调到最大 |
- |
buffer apply |
buffer apply |
使Buffer配置生效 |
- |
Spine 1 |
Spine 2 |
命令说明 |
注意事项 |
qos wred queue table 100G-WRED-Template |
qos wred queue table 100G-WRED-Template |
配置WRED表 |
- |
queue 0 drop-level 0 low-limit 37999 high-limit 38000 discard-probability 20 queue 0 drop-level 1 low-limit 37999 high-limit 38000 discard-probability 20 queue 0 drop-level 2 low-limit 37999 high-limit 38000 discard-probability 20 |
queue 0 drop-level 0 low-limit 37999 high-limit 38000 discard-probability 20 queue 0 drop-level 1 low-limit 37999 high-limit 38000 discard-probability 20 queue 0 drop-level 2 low-limit 37999 high-limit 38000 discard-probability 20 |
配置基于队列0的WRED表的内容 |
- |
queue 1 drop-level 0 low-limit 37999 high-limit 38000 discard-probability 20 queue 1 drop-level 1 low-limit 37999 high-limit 38000 discard-probability 20 queue 1 drop-level 2 low-limit 37999 high-limit 38000 discard-probability 20 |
queue 1 drop-level 0 low-limit 37999 high-limit 38000 discard-probability 20 queue 1 drop-level 1 low-limit 37999 high-limit 38000 discard-probability 20 queue 1 drop-level 2 low-limit 37999 high-limit 38000 discard-probability 20 |
配置基于队列1的WRED表的内容 |
- |
queue 2 drop-level 0 low-limit 37999 high-limit 38000 discard-probability 20 queue 2 drop-level 1 low-limit 37999 high-limit 38000 discard-probability 20 queue 2 drop-level 2 low-limit 37999 high-limit 38000 discard-probability 20 |
queue 2 drop-level 0 low-limit 37999 high-limit 38000 discard-probability 20 queue 2 drop-level 1 low-limit 37999 high-limit 38000 discard-probability 20 queue 2 drop-level 2 low-limit 37999 high-limit 38000 discard-probability 20 |
配置基于队列2的WRED表的内容 |
- |
queue 3 drop-level 0 low-limit 1000 high-limit 2000 discard-probability 20 queue 3 drop-level 1 low-limit 1000 high-limit 2000 discard-probability 20 queue 3 drop-level 2 low-limit 1000 high-limit 2000 discard-probability 20 queue 3 weighting-constant 0 queue 3 ecn |
queue 3 drop-level 0 low-limit 1000 high-limit 2000 discard-probability 20 queue 3 drop-level 1 low-limit 1000 high-limit 2000 discard-probability 20 queue 3 drop-level 2 low-limit 1000 high-limit 2000 discard-probability 20 queue 3 weighting-constant 0 queue 3 ecn |
配置基于队列3的WRED表的内容 |
RoCE队列的low-limit和high-limit配置较小值,其他队列配置较大值 |
queue 4 drop-level 0 low-limit 37999 high-limit 38000 discard-probability 20 queue 4 drop-level 1 low-limit 37999 high-limit 38000 discard-probability 20 queue 4 drop-level 2 low-limit 37999 high-limit 38000 discard-probability 20 |
queue 4 drop-level 0 low-limit 37999 high-limit 38000 discard-probability 20 queue 4 drop-level 1 low-limit 37999 high-limit 38000 discard-probability 20 queue 4 drop-level 2 low-limit 37999 high-limit 38000 discard-probability 20 |
配置基于队列4的WRED表的内容 |
- |
queue 5 drop-level 0 low-limit 37999 high-limit 38000 discard-probability 20 queue 5 drop-level 1 low-limit 37999 high-limit 38000 discard-probability 20 queue 5 drop-level 2 low-limit 37999 high-limit 38000 discard-probability 20 |
queue 5 drop-level 0 low-limit 37999 high-limit 38000 discard-probability 20 queue 5 drop-level 1 low-limit 37999 high-limit 38000 discard-probability 20 queue 5 drop-level 2 low-limit 37999 high-limit 38000 discard-probability 20 |
配置基于队列5的WRED表的内容 |
- |
queue 6 drop-level 0 low-limit 37999 high-limit 38000 discard-probability 20 queue 6 drop-level 1 low-limit 37999 high-limit 38000 discard-probability 20 queue 6 drop-level 2 low-limit 37999 high-limit 38000 discard-probability 20 queue 6 ecn |
queue 6 drop-level 0 low-limit 37999 high-limit 38000 discard-probability 20 queue 6 drop-level 1 low-limit 37999 high-limit 38000 discard-probability 20 queue 6 drop-level 2 low-limit 37999 high-limit 38000 discard-probability 20 queue 6 ecn |
配置基于队列6的WRED表的内容 |
- |
queue 7 drop-level 0 low-limit 37999 high-limit 38000 discard-probability 20 queue 7 drop-level 1 low-limit 37999 high-limit 38000 discard-probability 20 queue 7 drop-level 2 low-limit 37999 high-limit 38000 discard-probability 20 |
queue 7 drop-level 0 low-limit 37999 high-limit 38000 discard-probability 20 queue 7 drop-level 1 low-limit 37999 high-limit 38000 discard-probability 20 queue 7 drop-level 2 low-limit 37999 high-limit 38000 discard-probability 20 |
配置基于队列7的WRED表的内容 |
- |
Spine 1 |
Spine 2 |
命令说明 |
配置目的 |
interface HundredGigE1/1/1 |
interface HundredGigE1/1/1 |
进入接口视图 |
配置与Leaf互联链路 |
port link-mode route |
port link-mode route |
将以太网接口切换到三层工作模式 |
- |
ip address 172.16.2.153 255.255.255.252 |
ip address 172.16.3.153 255.255.255.252 |
配置IP |
- |
interface HundredGigE1/1/2 |
interface HundredGigE1/1/2 |
进入接口视图 |
配置与Leaf互联链路 |
port link-mode route |
port link-mode route |
将以太网接口切换到三层工作模式 |
- |
ip address 172.16.2.157 255.255.255.252 |
ip address 172.16.3.157 255.255.255.252 |
配置IP |
- |
interface HundredGigE1/1/3 |
interface HundredGigE1/1/3 |
进入接口视图 |
配置与Leaf互联链路 |
port link-mode route |
port link-mode route |
将以太网接口切换到三层工作模式 |
- |
ip address 172.16.2.81 255.255.255.252 |
ip address 172.16.3.81 255.255.255.252 |
配置IP |
- |
interface HundredGigE1/1/4 |
interface HundredGigE1/1/4 |
进入接口视图 |
配置与Leaf互联链路 |
port link-mode route |
port link-mode route |
将以太网接口切换到三层工作模式 |
- |
ip address 172.16.2.85 255.255.255.252 |
ip address 172.16.3.85 255.255.255.252 |
配置IP |
- |
Spine 1 |
Spine 2 |
命令说明 |
注意事项 |
interface range HundredGigE1/1/1 to HundredGigE1/1/4 |
interface range HundredGigE1/1/1 to HundredGigE1/1/4 |
配置与Leaf互联接口 |
- |
priority-flow-control deadlock enable |
priority-flow-control deadlock enable |
开启端口PFC死锁检测功能 |
- |
priority-flow-control enable |
priority-flow-control enable |
开启端口PFC功能 |
- |
priority-flow-control no-drop dot1p 3 |
priority-flow-control no-drop dot1p 3 |
开启RoCE报文所在队列的PFC功能 |
- |
priority-flow-control dot1p 3 headroom 491 |
priority-flow-control dot1p 3 headroom 491 |
配置3队列headroom缓存门限为491 |
|
priority-flow-control dot1p 3 reserved-buffer 20 |
priority-flow-control dot1p 3 reserved-buffer 20 |
配置PFC预留门限 |
|
priority-flow-control dot1p 3 ingress-buffer dynamic 5 |
priority-flow-control dot1p 3 ingress-buffer dynamic 5 |
配置动态反压帧触发门限 |
|
priority-flow-control dot1p 3 ingress-threshold-offset 12 |
priority-flow-control dot1p 3 ingress-threshold-offset 12 |
配置反压帧停止门限与触发门限间的偏移量 |
|
qos trust dscp |
qos trust dscp |
端口QoS信任DSCP |
- |
qos wfq byte-count |
qos wfq byte-count |
设置wfq队列调度 |
- |
qos wfq be group 1 byte-count 15 |
qos wfq be group 1 byte-count 15 |
设置0队列调度权重 |
根据实际业务情况配置调度权重 |
qos wfq af1 group 1 byte-count 2 |
qos wfq af1 group 1 byte-count 2 |
设置1队列调度权重 |
根据实际业务情况配置调度权重 |
qos wfq af2 group 1 byte-count 2 |
qos wfq af2 group 1 byte-count 2 |
设置2队列调度权重 |
根据实际业务情况配置调度权重 |
qos wfq af3 group 1 byte-count 60 |
qos wfq af3 group 1 byte-count 60 |
设置3队列调度权重 |
根据实际业务情况配置调度权重 本文以配置无损队列和其他队列权重占比为4:1为例 |
qos wfq cs6 group sp |
qos wfq cs6 group sp |
设置6队列调度优先级为sp |
根据实际业务情况配置调度权重 |
qos wfq cs7 group sp |
qos wfq cs7 group sp |
设置7队列调度优先级为sp |
根据实际业务情况配置调度权重 |
qos wred apply 100G-WRED-Template |
qos wred apply 100G-WRED-Template |
端口使能100G的WRED表 |
- |
qos gts queue 6 cir 50000000 cbs 16000000 |
qos gts queue 6 cir 50000000 cbs 16000000 |
CNP队列限速50G |
- |
Spine 1 |
Spine 2 |
命令说明 |
注意事项 |
interface range HundredGigE1/1/1 to HundredGigE1/1/4 |
interface range HundredGigE1/1/1 to HundredGigE1/1/4 |
配置与Spine 1互联接口 |
- |
priority-flow-control deadlock enable |
priority-flow-control deadlock enable |
开启端口PFC死锁检测功能 |
- |
priority-flow-control enable |
priority-flow-control enable |
开启端口PFC功能 |
- |
priority-flow-control no-drop dot1p 3 |
priority-flow-control no-drop dot1p 3 |
开启RoCE报文所在队列的PFC功能 |
- |
priority-flow-control dot1p 3 headroom 491 |
priority-flow-control dot1p 3 headroom 491 |
配置3队列headroom缓存门限为491 |
|
priority-flow-control dot1p 3 reserved-buffer 20 |
priority-flow-control dot1p 3 reserved-buffer 20 |
配置PFC预留门限 |
|
priority-flow-control dot1p 3 ingress-buffer dynamic 5 |
priority-flow-control dot1p 3 ingress-buffer dynamic 5 |
配置动态反压帧触发门限 |
|
priority-flow-control dot1p 3 ingress-threshold-offset 12 |
priority-flow-control dot1p 3 ingress-threshold-offset 12 |
配置反压帧停止门限与触发门限间的偏移量 |
|
qos trust dscp |
qos trust dscp |
端口QoS信任DSCP |
- |
qos wfq byte-count |
qos wfq byte-count |
设置WFQ队列调度 |
- |
qos wfq be group 1 byte-count 15 |
qos wfq be group 1 byte-count 15 |
设置0队列调度权重 |
根据实际业务情况配置调度权重 |
qos wfq af1 group 1 byte-count 2 |
qos wfq af1 group 1 byte-count 2 |
设置1队列调度权重 |
根据实际业务情况配置调度权重 |
qos wfq af2 group 1 byte-count 2 |
qos wfq af2 group 1 byte-count 2 |
设置2队列调度权重 |
根据实际业务情况配置调度权重 |
qos wfq af3 group 1 byte-count 60 |
qos wfq af3 group 1 byte-count 60 |
设置3队列调度权重 |
根据实际业务情况配置调度权重 本文以配置无损队列和其他队列权重占比为4:1为例 |
qos wfq cs6 group sp |
qos wfq cs6 group sp |
设置6队列调度优先级为sp |
根据实际业务情况配置调度权重 |
qos wfq cs7 group sp |
qos wfq cs7 group sp |
设置7队列调度优先级为sp |
根据实际业务情况配置调度权重 |
qos wred queue 3 drop-level 0 low-limit 1000 high-limit 2000 discard-probability 20 qos wred queue 3 drop-level 1 low-limit 1000 high-limit 2000 discard-probability 20 qos wred queue 3 drop-level 2 low-limit 1000 high-limit 2000 discard-probability 20 |
qos wred queue 3 drop-level 0 low-limit 1000 high-limit 2000 discard-probability 20 qos wred queue 3 drop-level 1 low-limit 1000 high-limit 2000 discard-probability 20 qos wred queue 3 drop-level 2 low-limit 1000 high-limit 2000 discard-probability 20 |
设置RoCE队列的ECN水线值 |
RoCE队列的low-limit和high-limit配置较小值,其他队列配置较大值 |
qos wred queue 3 weighting-constant 0 |
qos wred queue 3 weighting-constant 0 |
设置RoCE队列平均队列长度指数 |
- |
qos wred queue 3 ecn |
qos wred queue 3 ecn |
开启RoCE队列ecn |
- |
qos wred queue 6 drop-level 0 low-limit 37999 high-limit 38000 discard-probability 20 qos wred queue 6 drop-level 1 low-limit 37999 high-limit 38000 discard-probability 20 qos wred queue 6 drop-level 2 low-limit 37999 high-limit 38000 discard-probability 20 |
qos wred queue 6 drop-level 0 low-limit 37999 high-limit 38000 discard-probability 20 qos wred queue 6 drop-level 1 low-limit 37999 high-limit 38000 discard-probability 20 qos wred queue 6 drop-level 2 low-limit 37999 high-limit 38000 discard-probability 20 |
设置CNP队列的ECN水线值 |
- |
qos wred queue 6 ecn |
qos wred queue 6 ecn |
开启CNP队列ECN |
- |
qos gts queue 6 cir 50000000 cbs 16000000 |
qos gts queue 6 cir 50000000 cbs 16000000 |
CNP队列限速50G |
- |
Spine设备配置路由策略,将从Leaf收到的路由的AS替换为本地AS,否则Leaf设备无法学习其他Leaf发布的路由(设备收到BGP邻居发布的路由会检查AS_PATH属性,如果路由的AS_PATH中包含自身AS号,则不会学习该路由)。
Spine1 |
Spine2 |
命令说明 |
注意事项 |
ip as-path leaf_aspath permit 64636 |
ip as-path leaf_aspath permit 64636 |
配置一个AS路径过滤列表 |
- |
route-policy leaf_aspath_out permit node 10 |
route-policy leaf_aspath_out permit node 10 |
配置路由策略leaf_aspath_out的node 10 |
- |
if-match as-path leaf_aspath |
if-match as-path leaf_aspath |
通过AS路径过滤列表匹配路由 |
- |
apply as-path 64637 replace |
apply as-path 64637 replace |
配置BGP路由信息的AS_PATH属性,替换原有AS号 |
- |
route-policy leaf_aspath_out permit node 1000 |
route-policy leaf_aspath_out permit node 1000 |
配置路由策略leaf_aspath_out的node 1000 |
- |
Spine1 |
Spine2 |
命令说明 |
注意事项 |
bgp 64637 |
bgp 64637 |
启动BGP实例 |
- |
router-id 50.50.255.1 |
router-id 50.50.255.2 |
为指定的BGP实例配置BGP路由器的Router ID |
如果要在BGP实例下运行BGP协议,则必须为BGP实例指定Router ID, 如果没有指定Router ID,则缺省使用全局Router ID |
group Leaf external |
group Leaf external |
创建EBGP对等体组 |
- |
peer Leaf as-number 64636 |
peer Leaf as-number 64636 |
指定对等体组的AS号 |
- |
peer 172.16.2.154 group Leaf |
peer 172.16.3.154 group Leaf |
向对等体组中添加指定的对等体,即Leaf |
- |
peer 172.16.2.158 group Leaf |
peer 172.16.3.158 group Leaf |
向对等体组中添加指定的对等体,即Leaf |
- |
peer 172.16.2.82 group Leaf |
peer 172.16.3.82 group Leaf |
向对等体组中添加指定的对等体,即Leaf |
- |
peer 172.16.2.86 group Leaf |
peer 172.16.3.86 group Leaf |
向对等体组中添加指定的对等体,即Leaf |
- |
address-family ipv4 unicast |
address-family ipv4 unicast |
进入IPv4地址族 |
- |
balance 32 |
balance 32 |
配置BGP负载分担的路由条数 |
- |
peer Leaf enable |
peer Leaf enable |
允许本地路由器与指定对等体交换IPv4单播路由信息 |
- |
peer Leaf route-policy leaf_aspath_out export |
peer Leaf route-policy leaf_aspath_out export |
为IPv4对等体设置基于路由策略的路由修改下一跳策略 |
- |
修改Leaf或Spine数据缓冲区和PFC的配置可能引起丢包,例如实验室环境测试S6850设备R6710版本丢包情况如表1-1所示,现网环境丢包情况与组网规模和配置有关,请以具体情况为准。
命令 |
命令说明 |
丢包情况 |
buffer apply |
使数据缓冲区的配置生效 |
单条流丢480ms,整机丢秒级 |
priority-flow-control no-drop dot1p 3 |
使能/去使能端口PFC优先级 |
丢包30ms左右 |
priority-flow-control dot1p 5 headroom 125 |
修改PFC的headroom值 |
丢包10ms内 |
priority-flow-control dot1p 5 reserved-buffer 17 |
修改PFC的reserved-buffer值 |
丢包10ms内 |
priority-flow-control dot1p 5 ingress-buffer dynamic 5 |
修改PFC入方向门限dynamic值 |
丢包10ms内 |
priority-flow-control dot1p 5 ingress-threshold-offset 12 |
修改PFC入方向反压帧停止门限与触发门限间的偏移量 |
丢包10ms内 |
undo stp edged-port |
端口视图取消端口为边缘端口 |
对应接口学习的mac会删除,二层流会广播 |
qos wfq byte-count |
端口配置WFQ队列调度 |
丢包约40ms |
qos wfq af3 group 1 byte-count 50 |
端口配置3队列调度权重 |
丢包约30ms |
qos map-table import 1 export 1 |
修改设备qos-map对应表 |
不丢包 |
qos wred apply 100G-WRED-Template |
端口应用WRED模板 |
不丢包 |
queue 3 drop-level 1 low-limit 37999 high-limit 38000 discard-probability 20 |
修改ECN的水线值及丢弃概率 |
不丢包 |
queue 5 weighting-constant 0 |
修改WRED计算平均队列长度的指数 |
不丢包 |
qos gts queue 3 cir 4000000 cbs 16000000 |
配置GTS限速 |
不丢包 |
stp edged-port |
配置端口为边缘端口 |
不丢包 |
priority-flow-control enable |
端口使能PFC |
不丢包 |
priority-flow-control deadlock enable |
端口配置PFC死锁 |
不丢包 |
buffer threshold alarm |
全局配置buffer告警 |
不丢包 |
配置MQC,动作为accounting |
配置MQC,动作为accounting |
不丢包 |
配置端口入方向MQC,匹配dot1p,remark dscp |
端口下发入方向MQC |
不丢包 |
配置端口出方向MQC,匹配dscp,remark dot1p |
端口下发出方向MQC |
不丢包 |
priority-flow-control poolID 0 headroom 9928 |
全局配置PFC的headroom值 |
不丢包 |
dldp enable |
端口开启DLDP功能 |
不丢包 |
unicast-suppression 1 |
端口使能/去使能未知单播抑制 |
不丢包 |
broadcast-suppression |
端口使能/去使能广播抑制 |
不丢包 |
multicast-suppression |
端口使能/去使能组播抑制 |
不丢包 |
stp root-protection |
端口开启/关闭根保护功能 |
不丢包 |
grpc相关配置 |
全局配置GRPC相关配置 |
不丢包 |
编号 |
类型 |
流量方向 |
流量路径 |
仿真方式 |
流量大小 |
说明 |
1 |
单播/L2 |
同组M-LAG Leaf东西 |
M-LAG接入:Server1-Leaf1/Leaf2-Server2 单挂:Server3-Leaf3- Leaf4-Server4 |
测试仪 |
轻载/拥塞 |
Server1与server2之间二层互通; |
2 |
单播/L2 |
同组M-LAG Leaf东西 |
M-LAG接入:Server2-Leaf2/Leaf1-Server1 单挂:Server4-Leaf4 –Leaf3-Server3 |
测试仪 |
轻载/拥塞 |
Server2与server1之间二层互通; |
3 |
已知单播/IPV4 |
同组M-LAG Leaf东西 |
M-LAG接入:Server1-Leaf1/Leaf2-Server2 单挂:Server3-Leaf3 –Leaf4-Server4 |
测试仪 |
轻载/拥塞 |
Server1与server2之间三层互通; |
4 |
已知单播/IPV4 |
同组M-LAG Leaf东西 |
M-LAG接入:Server2-Leaf2/Leaf1-Server1 单挂:Server4-Leaf4 –Leaf3-Server3 |
测试仪 |
轻载/拥塞 |
Server2与server1之间三层互通; |
5 |
已知单播/IPV4 |
不同组M-LAG Leaf东西 |
Server1-Leaf1/Leaf2-Spine 1/Spine 2-Leaf3-Server3 |
测试仪 |
轻载/拥塞 |
Server1与server3之间三层互通; |
6 |
已知单播/IPV4 |
不同组M-LAG Leaf东西 |
Server3-Leaf3-Spine 1/Spine 2-Leaf1/Leaf2-Server1 |
测试仪 |
轻载/拥塞 |
Server1与server3之间三层互通; |
表1-2 链路故障测试情况表
设备 |
故障类型 |
流量中断时间 |
故障类型 |
流量中断时间 |
Leaf |
单上行链路故障 |
420ms |
单上行链路恢复 |
0ms |
Leaf双上行链路故障 |
330ms |
Leaf双上行链路恢复 |
0ms |
|
下行单挂故障 |
120ms |
下行单挂恢复 |
0ms |
|
peer-link链路故障(备设备被mad down) |
740ms |
peer-link链路故障恢复 |
0ms |
|
peer-link链路成员口故障 |
10ms |
peer-link链路成员口恢复 |
0ms |
|
Keepalive链路故障 |
0ms |
Keepalive链路故障恢复 |
0ms |
|
M-LAG主设备整机重启(重启时) |
170ms |
M-LAG主设备整机重启(恢复时) |
0ms(若重新选举则丢包) |
|
M-LAG备设备整机重启(重启时) |
170ms |
M-LAG备设备整机重启(恢复时) |
0ms |
|
Spine |
整机重启(重启时) |
30ms |
整机重启(恢复时) |
0ms |
下行ECMP链路单点故障 |
160ms |
下行ECMP链路单点故障恢复 |
0ms |
表1-3 配置验证命令汇总表
Leaf 1 |
Leaf 2 |
命令说明 |
display m-lag summary |
display m-lag summary |
显示M-LAG系统的接口摘要信息 |
display m-lag system |
display m-lag system |
显示M-LAG系统信息 |
display m-lag keepalive |
display m-lag keepalive |
显示M-LAG Keepalive报文的信息 |
display m-lag role |
display m-lag role |
显示M-LAG设备角色信息 |
display m-lag consistency-check status |
display m-lag consistency-check status |
显示M-LAG配置一致性检查状态 |
display vrrp verbose |
display vrrp verbose |
查看VRRP信息 |
display priority-flow-control interface |
display priority-flow-control interface |
显示接口的PFC信息 |
display qos wred interface |
display qos wred interface |
查看接口的ECN信息 |
display qos map-table dot1p-lp |
display qos map-table dot1p-lp |
查看dot1p-lp的映射关系 |
display qos map-table dscp-dot1p |
display qos map-table dscp-dot1p |
查看dscp-dot1p的映射关系 |
# 在Leaf1上查看Leaf 1与Leaf 2之间的M-LAG系统状态,M-LAG正常建立。
[leaf1] display m-lag summary
Flags: A -- Aggregate interface down, B -- No peer M-LAG interface configured
C -- Configuration consistency check failed
Peer-link interface: BAGG1000
Peer-link interface state (cause): UP
Keepalive link state (cause): UP
M-LAG interface information
M-LAG IF M-LAG group Local state (cause) Peer state Remaining down time(s)
BAGG100 1 UP UP -
# 在Leaf 1上查看M-LAG系统信息。
[leaf1] display m-lag system
System information
Local system number: 1 Peer system number: 2
Local system MAC: 2001-0000-0018 Peer system MAC: 2001-0000-0018
Local system priority: 110 Peer system priority: 110
Local bridge MAC: 0068-5716-5701 Peer bridge MAC: 90e7-10b2-f8aa
Local effective role: Primary Peer effective role: Secondary
Health level: 0
Standalone mode on split: Disabled
In standalone mode: No
System timer information
Timer State Value (s) Remaining time (s)
Auto recovery Disabled - -
Restore delay Disabled 30 -
Consistency-check delay Disabled 15 -
Standalone delay Disabled - -
Role to None delay Disabled 60 -
# 在Leaf 1上查看M-LAG Keepalive报文的信息。
[leaf1] display m-lag keepalive
Neighbor keepalive link status (cause): Up
Neighbor is alive for: 451077 s 86 ms
Keepalive packet transmission status:
Sent: Successful
Received: Successful
Last received keepalive packet information:
Source IP address: 1.1.1.2
Time: 2022/06/07 16:19:43
Action: Accept
M-LAG keepalive parameters:
Destination IP address: 1.1.1.2
Source IP address: 1.1.1.1
Keepalive UDP port : 6400
Keepalive interval : 1000 ms
Keepalive timeout : 5 sec
Keepalive hold time: 3 sec
# 在Leaf 1上查看M-LAG设备角色信息。
[leaf1] display m-lag role
Effective role information
Factors Local Peer
Effective role Primary Secondary
Initial role None None
MAD DOWN state Yes Yes
Health level 0 0
Role priority 50 100
Bridge MAC 0068-5716-5701 90e7-10b2-f8aa
Effective role trigger: Peer link calculation
Effective role reason: Role priority
Configured role information
Factors Local Peer
Configured role Primary Secondary
Role priority 50 100
Bridge MAC 0068-5716-5701 90e7-10b2-f8aa
# 在Leaf 1上查看M-LAG配置一致性信息。
[leaf1] display m-lag consistency-check status
Global Consistency Check Configuration
Local status : Enabled Peer status : Enabled
Local check mode : Strict Peer check mode : Strict
Consistency Check on Modules
Module Type1 Type2
LAGG Check Check
VLAN Check Check
STP Check Check
MAC Not Check Check
L2VPN Not Check Check
PORTSEC Not Check Not Check
DOT1X Not Check Not Check
MACA Not Check Not Check
WEBAUTH Not Check Not Check
NETANALYSIS Not Check Check
Type1 Consistency Check Result
Global consistency check result: SUCCESS
Inconsistent global modules: -
M-LAG IF M-LAG group ID Check Result Inconsistency modules
BAGG100 1 SUCCESS -
通常情况下,建议参考最佳实践/配置指导书组网,并采用PFC和ECN的推荐配置值。推荐值在一般的组网环境下是效果较好的参数组合,一般不需要调整。
有些情况下,由于组网需求或网络中的设备、服务器存在特殊性,需要进一步调整参数时,请参考1.12 参数调整指导。
对于S6805 & S6825 & S6850 & S9850、S9820-64H、S9820-8C、S9825&S9855设备,开启指定802.1p优先级的PFC功能后,设备会为PFC的各种门限设置一个缺省值,具体请参考表1-4、表1-5、表1-6。
此缺省值在一般的组网环境下是效果较好的参数组合,一般不建议调整。
表1-4 S6805 & S6825 & S6850 & S9850 PFC门限缺省值
PFC门限(右) 接口类型(下) |
Headroom缓存门限(cell) |
动态反压帧触发门限(%) |
反压帧停止门限与触发门限间的偏移量(cell) |
PFC预留门限(cell) |
1GE/10GE |
100 |
5 |
12 |
17 |
25GE |
125 |
5 |
12 |
17 |
40GE |
200 |
5 |
12 |
17 |
50GE |
308 |
5 |
12 |
17 |
100GE |
491 |
5 |
12 |
17 |
表1-5 S9820-64H PFC门限缺省值
PFC门限(右) 接口类型(下) |
Headroom缓存门限(cell) |
动态反压帧触发门限(%) |
反压帧停止门限与触发门限间的偏移量(cell) |
PFC预留门限(cell) |
25GE |
125 |
5 |
12 |
20 |
100GE |
491 |
5 |
12 |
20 |
表1-6 S9820-8C PFC门限缺省值
PFC门限(右) 接口类型(下) |
Headroom缓存门限(cell) |
动态反压帧触发门限(%) |
反压帧停止门限与触发门限间的偏移量(cell) |
PFC预留门限(cell) |
100GE |
491 |
5 |
12 |
20 |
200GE |
705 |
5 |
12 |
20 |
400GE |
1000 |
5 |
12 |
20 |
对于ECN功能,RoCE队列推荐使用表1-7的配置值,其他队列可按需配置较大的平均长度下限值和平均长度上限值。
表1-7 ECN各门限推荐值
PFC门限(右) 接口类型(下) |
队列平均长度的下限 |
队列平均长度的上限 |
丢弃概率 |
计算平均队列长度的指数 |
25GE |
400 |
1625 |
20 |
0 |
50GE |
600 |
1500 |
20 |
0 |
100GE |
1000 |
2000 |
20 |
0 |
200GE |
1500 |
3000 |
20 |
0 |
400GE |
2100 |
5000 |
20 |
0 |
确认服务器网卡是否支持和启用了PFC和ECN功能(具体操作方式请参考服务器的资料说明),并查看服务器的RoCE队列报文是否携带ECN标志位。
本文以Mellanox ConnectX-6 Lx网卡的配置作为示例。
需要注意的是:网卡配置在服务器重启和网卡重启后会失效,重启服务器或重启网卡后需要重新配置(网卡配置没有写到配置文件中,因此重启会失效)。
项目 |
信息 |
网卡型号 |
Ethernet controller: Mellanox Technologies MT2894 Family [ConnectX-6 Lx] |
网卡驱动版本 |
MLNX_OFED_LINUX-5.4-3.2.7.2.3-rhel8.4-x86_64 |
网卡固件版本 |
driver: mlx5_core version: 5.4-3.2.7.2.3 firmware-version: 26.31.2006 (MT_0000000531) |
执行mst start命令开启MST(Mellanox Software Tools)服务。
执行mst status查看Mellanox设备状态信息。
(可选)执行show_gids命令查看网卡名称、GID值、IP地址。
执行mlnx_qos -i ifname --trust dscp命令配置接口信任报文的DSCP优先级,ifname为网卡接口名称。
另外需要注意,网卡上设置的报文的DSCP值需要与设备上RoCE报文的DSCP优先级对应,即设备上开启PFC和ECN功能的802.1p优先级映射的DSCP优先级等于网卡发出报文的DSCP优先级。
步骤如下:
(1) 使用ethtool -i ifname bus-info命令查看接口的bus-info值。例如查看ens1f0接口的bus-info值执行ethtool -i ens1f0 bus-info,查询到ens1f0接口的bus-info为0000:86:00.0。
(2) 进入设置DSCP优先级的路径cd /sys/kernel/debug/mlx5/0000:86:00.0/cc_params
(3) 执行echo priority_value > np_cnp_dscp命令设置CNP报文的DSCP优先级,例如设置CNP报文的DSCP优先级为48执行echo 48 > np_cnp_dscp。
(4) 执行cat np_cnp_dscp命令查看CNP报文优先级是否修改成功。
执行mlnx_qos -i ifname ––pfc 0,1,2,3,4,5,6,7命令开启RoCE队列的PFC功能。ifname为网卡接口名称。0,1,2,3,4,5,6,7对应位置的值设置为0表示数值代表的优先级不开启PFC,设置为1表示数值代表的优先级开启PFC。
例如,开启接口ens1f0 802.1P优先级为5的报文的PFC命令为mlnx_qos -i ens1f0 –-pfc 0,0,0,0,0,1,0,0。
执行mlnx_qos -i ifname命令可以查看接口PFC功能开启情况。值为1表示该优先级的报文开启PFC。
执行如下命令开启指定优先级报文的ECN功能:
echo 1 > /sys/class/net/ifname/ecn/roce_np/enable/priority_value
echo 1 > /sys/class/net/ifname/ecn/roce_rp/enable/priority_value
例如开启接口ens1f0优先级为5的报文的ECN功能执行:
echo 1 > /sys/class/net/ens1f0/ecn/roce_np/enable/5
echo 1 > /sys/class/net/ens1f0/ecn/roce_rp/enable/5
执行下面命令查看ECN是否开启成功,返回1表示开启成功,返回为0表示没有开启。
cat /sys/class/net/ifname/ecn/roce_np/enable/priority_value
cat /sys/class/net/ifname/ecn/roce_rp/enable/priority_value
在进行参数调整的过程中,配置如下PFC、QoS、数据缓冲区相关命令时会导致端口流量中断。
· buffer apply
· buffer egress cell queue shared(配置本命令本身不丢包,执行buffer apply应用配置时丢包)
· qos wred apply
· qos wrr weight
· qos wrr group weight
· qos wfq byte-count
· qos wfq queue-id group { 1 | 2 } byte-count
· priority-flow-control no-drop dot1p
· priority-flow-control dot1p headroom
· priority-flow-control dot1p ingress-buffer dynamic
· priority-flow-control dot1p ingress-buffer static
· priority-flow-control dot1p ingress-threshold-offset
· priority-flow-control dot1p reserved-buffer
查看接口HundredGigE1/0/25丢弃报文的信息。
<Sysname> display packet-drop interface hundredgige 1/0/25
HundredGigE1/0/25:
Packets dropped due to Fast Filter Processor (FFP): 0
Packets dropped due to Egress Filter Processor (EFP): : 0
Packets dropped due to STP non-forwarding state: 0
Packets dropped due to insufficient data buffer. Input dropped: 0 Output dropped: 0
Packets of ECN marked: 0
Packets of WRED droped: 0
(1) 若出现Input dropped类型的丢包,可以增加端口Headroom缓存门限值的大小,建议按照当前配置值的倍数进行调整,直到不再出现Input dropped类型的丢包。例如:
# 配置Headroom缓存门限为982。
<sysname> system-view
[Sysname] interface hundredgige 1/0/25
[Sysname-HundredGigE1/0/25] priority-flow-control dot1p 5 headroom 982
如果调整至最大值仍然丢包,请排查服务器网卡是否支持和启用PFC功能。
(2) 若出现Output dropped类型的丢包,需要检查队列出方向最大共享区域占用比是否配置为100%和是否执行了buffer apply命令。例如:
# 配置队列5在cell资源中的最大共享缓存占用比为100%并应用配置。
<sysname> system-view
[Sysname] buffer egress cell queue 5 shared ratio 100
[Sysname] buffer apply
如果出方向队列配置为100%后仍然有Output dropped类型的丢包,则可能存在组网或配置问题,请排查组网和配置问题或联系技术支持。
(1) 若出现WRED droped丢包,请查看RoCE队列的ECN功能是否已开启。
<Sysname> display qos wred table
Table name: 1
Table type: Queue based WRED
QID gmin gmax gprob ymin ymax yprob rmin rmax rprob exponent ECN
----------------------------------------------------------------------------
0 100 1000 10 100 1000 10 100 1000 10 9 N
1 100 1000 10 100 1000 10 100 1000 10 9 N
2 100 1000 10 100 1000 10 100 1000 10 9 N
3 100 1000 10 100 1000 10 100 1000 10 9 N
4 100 1000 10 100 1000 10 100 1000 10 9 N
5 100 1000 10 100 1000 10 100 1000 10 9 N
6 100 1000 10 100 1000 10 100 1000 10 9 N
7 100 1000 10 100 1000 10 100 1000 10 9 N
(2) 若RoCE队列的ECN功能未开启(显示信息中ECN字段显示为N),则需要开启RoCE队列的ECN功能。
<Sysname> system-view
[Sysname] interface hundredgige 1/0/25
[Sysname-HundredGigE1/0/25] qos wred queue 5 ecn
(3) 若RoCE队列的ECN功能开启后WRED类型的丢包继续增加,则需要查看服务器的RoCE队列报文是否携带ECN标志位。若未携带,请排查服务器网卡是否启用了PFC和ECN功能。
(4) 如果仍然存在丢包,请联系技术支持。
为了满足时延需求,在不丢包的情况下,可以调整ECN和PFC的部分参数,进而降低时延。调整时延会影响吞吐,可以认为的原则是:在拥塞网络环境下,占用的buffer越小,时延越低,吞吐量较小;占用的buffer越大,时延越大,吞吐量较大。请根据具体组网需求平衡时延和吞吐的关系。
不拥塞网络环境下,PFC和ECN等功能不生效。触发ECN、PFC后,报文转发降速。调整的原则是尽量在低buffer占用,不丢包的情况下触发ECN和PFC。建议优先调整ECN。
另外需要注意的是,参数调整还需要考虑服务器网卡的支持情况,包括网卡对PFC、ECN报文的响应机制,是否有PFC、ECN相关的特殊能力。例如有些网卡支持自降速功能(时延大的情况自动降速)。此时如果我们想要加大吞吐,但是调整后加大了时延,服务器自降速了,反而起到了和预期相反的效果。建议调整前先查阅服务器的相关资料,了解服务器网卡PFC、ECN功能的支持情况和相关功能,并在调整过程中实际验证调整是否合适。
可以调整WRED表中各队列的如下参数,控制时延与吞吐量:
· Low-limit(队列平均长度的下限)和high-limit(队列平均长度的上限):通过调低low-limit和high-limit值,可以更快的触发ECN标记,使时延更低,但可能会影响吞吐量。
· Weighting-constan(计算平均队列长度的指数):表示的是队列深度的计算方式,为0表示实时队列深度,对ECN的标记会更灵敏。平均队列长度的指数越大,计算平均队列长度时对队列的实时变化越不敏感。建议保持推荐值。仅在完成其他调整项后仍然无效的情况下再尝试调整。
· discard-probability(丢弃概率):在使能了ECN后,代表的是标记概率,该值越大,表示在上下限门限之间被标记的报文数越多。建议保持推荐值,仅在完成其他调整项后仍然无效的情况下再尝试调整。可以按照当前值20%的幅度进行调整。
配置举例:
# 配置队列5的WRED参数:队列平均长度的下限为800,队列平均长度的上限为1800。
<Sysname> system-view
[Sysname] qos wred queue table queue-table1
[Sysname-wred-table-queue-table1] queue 5 drop-level 0 low-limit 800 high-limit 1800
[Sysname-wred-table-queue-table1] queue 5 drop-level 1 low-limit 800 high-limit 1800
[Sysname-wred-table-queue-table1] queue 5 drop-level 2 low-limit 800 high-limit 1800
PFC是ECN后的一层屏障,保证设备不丢包。PFC一般不被触发,对时延影响不大,并且PFC门限调整过低会影响吞吐量,所以一般不建议调整。
但如果想进一步降低时延,可以尝试调整ingress-buffer参数(反压帧触发门限)。
需要注意的是,在设备有流量收发的情况下调整PFC门限会引起丢包。
(1) 减小ingress-buffer门限值,建议跟随WRED的high-limit门限进行调整(按当前值10%的幅度减小),但要保证该门限值大于WRED的high-limit门限值,保证设备优先触发ECN功能。
# 配置动态反压帧触发门限为4。
<sysname> system-view
[Sysname] interface hundredgige 1/0/25
[Sysname-HundredGigE1/0/25] priority-flow-control dot1p 5 ingress-buffer dynamic 4
(2) 调整完成后,可以通过命令行多次查看接口的PFC信息,尽量保证设备上PFC的收发数量较少,即不触发或偶尔触发PFC。如果多次查看发现存在PFC报文收发,则表示ingress-buffer值配置过低,建议回调。
# 显示接口的PFC信息。
<Sysname> display priority-flow-control interface
Conf -- Configured mode Ne -- Negotiated mode P -- Priority
Interface Conf Ne Dot1pList P Recv Sent Inpps Outpps
HGE1/0/25 Auto On 0,2-3,5-6 0 178 43 12 15
升级前的命令行检查,请参见“1.10 配置验证命令行”和以下表格中的命令行。
表1-8 升级前的命令行检查汇总表
Leaf 1 |
Leaf 2 |
命令说明 |
display device |
display device |
用来显示设备信息 |
display boot-loader |
display boot-loader |
用来显示本次启动和下次启动所采用的启动软件包的名称 |
display version |
display version |
用来显示系统版本信息 |
具体操作步骤请参考“H3C交换机M-LAG升级、替换、扩容指导”。
单点故障(轻载)时间正切小于500ms,回切小于50ms(传统升级方式,两台M-LAG设备逐台升级)。具体参见“1.9 收敛时间测试”。
升级后的命令行检查,请参见“1.10 配置验证命令行”和以下表格中的命令行。
表1-9 升级后的命令行检查汇总表
Leaf 1 |
Leaf 2 |
命令说明 |
display device |
display device |
用来显示设备信息 |
display boot-loader |
display boot-loader |
用来显示本次启动和下次启动所采用的启动软件包的名称 |
display version |
display version |
用来显示系统版本信息 |
升级前的命令行检查,请参见“1.10 配置验证命令行”和以下表格中的命令行。
表1-10 升级前的命令行检查汇总表
Spine 1 |
Spine 2 |
命令说明 |
display device |
display device |
用来显示设备信息 |
display boot-loader |
display boot-loader |
用来显示本次启动和下次启动所采用的启动软件包的名称 |
display version |
display version |
用来显示系统版本信息 |
升级设备软件前,请进行如下操作:
(1) 使用display version命令查看设备当前运行的BootWare程序以及启动软件的版本。
(2) 获取新软件的版本发布说明书,了解新软件的版本号、软件大小以及和当前运行的BootWare程序以及Comware软件的兼容性。
(3) 通过版本发布说明书了解将安装的软件包是否需要License。如果需要,查看设备上是否有对应的有效的License。如果没有,请先安装License。否则,会导致软件包安装失败。
(4) 使用dir命令查看存储介质是否有足够的空间存储新的软件。如果存储空间不足,可使用delete命令删除一些暂时不用的文件。请保证系统中所有的主控板都有足够的存储空间。
(5) 使用FTP、TFTP方式将新软件包下载到任一文件系统的根目录下。
具体升级方式没有特殊要求,以现场情况为准。
单点故障(轻载)时间正切小于200ms,回切小于50ms。具体参见“1.9 收敛时间测试”。
升级后的命令行检查,请参见“1.10 配置验证命令行”和以下表格中的命令行。
表1-11 升级后的命令行检查汇总表
Leaf 1 |
Leaf 2 |
命令说明 |
display device |
display device |
用来显示设备信息 |
display boot-loader |
display boot-loader |
用来显示本次启动和下次启动所采用的启动软件包的名称 |
display version |
display version |
用来显示系统版本信息 |
本文描述的扩容操作指新增一组Leaf设备。
扩容前的命令行检查,请参见“1.10 配置验证命令行”和以下表格中的命令行。
表1-12 扩容前的命令行检查汇总表
Leaf 1 |
Leaf 2 |
命令说明 |
display device |
display device |
用来显示设备信息 |
display boot-loader |
display boot-loader |
用来显示本次启动和下次启动所采用的启动软件包的名称 |
display version |
display version |
用来显示系统版本信息 |
(1) 新增设备断开网管
(2) 升级完成相应版本
(3) 做好预配置
(4) 接入网管
扩容后的命令行检查,请参见下表中的命令行。
表1-13 扩容后的命令行检查汇总表
Leaf 1 |
Leaf 2 |
命令说明 |
display device |
display device |
用来显示设备信息 |
display boot-loader |
display boot-loader |
用来显示本次启动和下次启动所采用的启动软件包的名称 |
display version |
display version |
用来显示系统版本信息 |
替换前的命令行检查,请参见“1.10 配置验证命令行”和以下表格中的命令行。
表1-14 替换前的命令行检查汇总表
Leaf 1 |
Leaf 2 |
命令说明 |
display device |
display device |
用来显示设备信息 |
display boot-loader |
display boot-loader |
用来显示本次启动和下次启动所采用的启动软件包的名称 |
display version |
display version |
用来显示系统版本信息 |
相关业务板上的业务及网管流量应先行切换到正常业务板上。
支持业务板热插拔或者整机断电替换,具体根据现场情况而定。
具体操作步骤请参考“H3C交换机M-LAG升级、替换、扩容指导”。
与替换前命令行检查相同。
替换前的命令行检查,请参见“1.10 配置验证命令行”和以下表格中的命令行。
表1-15 替换前的命令行检查汇总表
Leaf 1 |
Leaf 2 |
命令说明 |
display device |
display device |
用来显示设备信息 |
display boot-loader |
display boot-loader |
用来显示本次启动和下次启动所采用的启动软件包的名称 |
display version |
display version |
用来显示系统版本信息 |
支持网板热插拔或者整机断电替换,具体根据现场情况而定。
与替换前命令行检查相同。
不同款型规格的资料略有差异, 详细信息请向具体销售和400咨询。H3C保留在没有任何通知或提示的情况下对资料内容进行修改的权利!