What Should I Do If the Remote Server Fails to Be Accessed Through the NAT Gateway?¶
If your TCP connection fails when your ECS is accessing a server on the public network through an SNAT rule, perform the following steps:
Run the following command to check whether tcp_tw_recycle is enabled on the remote server:
sysctl -a|grep tcp_tw_recycle
If the value of tcp_tw_recycle is 1, tcp_tw_recycle is enabled.
Run the following command to check the number of lost packets of the remote server:
cat /proc/net/netstat | awk '/TcpExt/ { print $21,$22 }'
If the value of ListenDrops is not 0, packet loss occurs, that is, the network is faulty.
Troubleshooting¶
Method 1: Modifying the kernel parameter of the remote server
Run the following command to temporarily modify the parameters (the modification becomes invalid after the server is restarted):
sysctl -w net.ipv4.tcp_tw_recycle=0
Perform the following operations to permanently modify the parameters:
Modify the /etc/sysctl.conf file:
vi /etc/sysctl.conf
Add the following content to the file:
net.ipv4.tcp_tw_recycle=0
Press Esc, enter :wq!, and save the file and exit.
Run the following command to make the modification take effect:
sysctl -p
Method 2: Modifying the kernel parameter of the local client
To temporarily modify parameters (the settings become invalid after the local client is restarted), configure the parameter as follows:
sysctl -w net.ipv4.tcp_timestamps=0
Perform the following operations to permanently modify the parameters:
Modify the /etc/sysctl.conf file:
vi /etc/sysctl.conf
Add the following content to the file:
net.ipv4.tcp_timestamps=0
Press Esc, enter :wq!, and save the file and exit.
Run the following command to make the modification take effect:
sysctl -p