具体型号

HUAWEI CE6865E-48S8CQ uptime is 0 day, 0 hour, 23 minutes 
Patch Version: V200R022SPH010

当前在A设备预先有的配置

assign forward eth-trunk mode 256
#
info-center loghost source MEth0/0/0
info-center loghost 192.168.248.65 vpn-instance OB
info-center loghost 192.168.248.66 vpn-instance OB
#
transceiver non-certified-alarm disable
#
system resource standard
#
device board 1 board-type CE6865E-48S8CQ
device board 2 board-type CE6865E-48S8CQ
#
drop-profile default
#               
dcb pfc
#
dcb ets-profile default
#
ntp server disable
ntp ipv6 server disable
ntp server source-interface all disable
ntp ipv6 server source-interface all disable
ntp unicast-server 192.168.8.102 vpn-instance OB
#
icmp name timestamp-reply send disable
#
telnet server disable
telnet ipv6 server disable
undo telnet server-source all-interface
undo telnet ipv6 server-source all-interface
#
diffserv domain default
#
ip vpn-instance DAD
 ipv4-family
  route-distinguisher 9:9
#               
ip vpn-instance OB
 ipv4-family
  route-distinguisher 11:1
#
dhcp server request-packet all-interface disable
......


堆叠A设备

stack
stack member 1 priority 150
stack member 1 domain 10
interface stack-port 1/1
port member-group interface 100ge 1/0/1 to 1/0/4
y
save
reboot
interface stack-port 1/1
shutdown

堆叠B设备

stack
stack member 1 priority 100
stack member 1 domain 10
stack member 1 renumber 2 inherit-config
interface stack-port 1/1
port member-group interface 100ge 1/0/1 to 1/0/4
y
save
reboot
在重启的同时进入A设备
interface stack-port 1/1
undo shutdown

配置直连双主检测

int 25ge 1/0/47
dual-active detect mode direct
int 25ge 1/0/48
dual-active detect mode direct
int 25ge 2/0/47
dual-active detect mode direct
int 25ge 2/0/48
dual-active detect mode direct

此时就发现问题了，接口指示灯根本不亮，查看接口状态显示ERROR DOWN(resource-mismatch)，资源状态不一致？？？？

然后我就想到了可能是这条命令 assign forward eth-trunk mode 256 导致的；
这条命令是设置聚合组数量的，设置后需要重启设备。
因为A设备在配置堆叠命令后 已经重启了，所以实际是已经生效了的。
然后B设备重启后才堆叠成功，A设备在将配置同步给B设备，此时assign forward eth-trunk mode 256才配上去，因为没有重启的原因，导致资源状态不一致，导致直连双主检测根本起不来。

解决方法，重启一下B设备，重启后直连双主检测起来了，接口指示灯正常发光。