首先，现场的场景是上联二层SW，这个时候上联SW上行口down掉之后，上联SW的下行接口不会down，导致Fw感知不到光猫的接口有问题，所以要配置track+nqa，并且在RBM视图下调用，一定要记得，先保证track的状态是postive的
然后再在RBM视图下引用
nqa entry admin rbm
type icmp-echo
destination ip x.x.x.x
frequency 100
out interface GigabitEthernet1/0/28
reaction 1 checked-element probe-fail threshold-type consecutive 5 action-type trigger-only
#
nqa entry admin rbm1
type icmp-echo
destination ip x.x.x.x
frequency 100
out interface GigabitEthernet1/0/29
reaction 1 checked-element probe-fail threshold-type consecutive 5 action-type trigger-only
nqa schedule admin rbm start-time now lifetime forever
nqa schedule admin rbm1 start-time now lifetime forever
track 4 nqa entry admin rbm reaction 1
track 5 nqa entry admin rbm1 reaction 1
remote-backup group
backup-mode dual-active
data-channel interface Route-Aggregation1
configuration sync-check interval 1
delay-time 1
track 4
track 5
local-ip 10.10.10.2
remote-ip 10.10.10.1
device-role primary
主设备上关于路由的配置要这样配置：
ip route-static 0.0.0.0 0 x.x.x.x track 4
ip route-static 0.0.0.0 0 x.x.x.x track 5 preference 70
其次，在两个出口下使用nat outbound 要使用地址池，并且要是对应的vrrp的虚地址
否则RBM主备切换，公网的流量回不来，并且对应地址池下要绑定VRRP的VID，否则会报地址冲突
nat address-group 1
address x.x.x.x x.x.x.x
vrrp vrid 1
#
nat address-group 2
address x.x.x.x x.x.x.x
vrrp vrid 2
interface GigabitEthernet1/0/28
port link-mode route
combo enable fiber
最后，要开启nat 链路转换重新创建会话，否则持续命中之前的会话，会有问题
nat link-switch recreate-session
#
补充一下，备机的RBM中不用track4 与track5
