No, this is correct, because I'm using this configuration and can guarantee normal connections
Just when I modify the interface, route will be reset
I need to manually execute the following command to connect to the network properly
route add 220.127.116.11 dev eth0
route add default gw 18.104.22.168 dev eth0
I am only saying that not being able to reach the gateway is the consequence of it being outside the defined network. If you insists in that configuration, I would try to add a "config 'route'" section, so it gets redefined each time the interface restarts.
Can you give an example?
In the morning, my friend used this method of configuration, but he failed
The above route statement is modified by the CentOS document of online.net
In fact, I don't understand its function
"route add 22.214.171.124 dev eth0" means: "host 126.96.36.199 can be found on eth0".
"route add default gw 188.8.131.52 dev eth0" means "all non-local traffic must be sent to host 184.108.40.206 using eth0".
Considering that you already configured 220.127.116.11 as the default gateway on your network config file, the second command seems redundant (and since you are also specifying the network device where 18.104.22.168 can be found, it also makes redundant the first command). To add that route in LEDE, I would add just these lines to the network config file: