Please give the relay server the ID check function.

[BlockShieldWeb3]

The current relay server selection logic of RustDesk is:
Client A connects to Relay Server A If Controller B wants to establish a connection with Client A, then Controller B must connect to Relay Server A. This relay server selection logic will lead to A and B being in a situation where there is a lot of latency in the network environment resulting in A and B connecting to the network. with a high latency in the network environment.
For this reason, we set up an intranet Relay Server IP:10.10.10.10
and mapped the Server to three different public IPs through SDWAN acceleration: IP1:ChinaTelecom IP2:ChinaUnicom IP3:GoogleAnyCastIP
We used private fiber links to connect to these three networks, and all three public IPs correspond to the same intranet Relay Server.
On the console configuration of RustDesk Pro, we set up the DNS smart resolution for relay.company.com. If Client A's network is ChinaTelecom, the DNS resolution for the domain name will result in IP1 , If Controller B's network is a Singapore carrier, then the DNS resolution result for the domain name is IP3. However, when we try to connect from Controller B in Singapore to Client A in Mainland China, we find that Client A connects to IP3 instead of IP1 according to the DNS resolution result, which results in high latency.
I think the relay server should add a relay server ID identification function, when the same relay server is published through different public IP, between the two ends that need to establish a connection, if the relay server is recognized as a different public IP but the relay server ID is the same, TCPING should be used to determine which public IP has a lower latency to establish a connection with which relay server, so that This can greatly optimize the connection delay due to network restrictions in different countries.

Thank the RustDesk Team!