Connection between the RTI and federates is closed after some idle time

[byeonggiljun]

When we connect a federate to the RTI on a different machine, it seems there is a timeout for the connection.

When I ran the example above (I didn't run the RTI on the local host), the federate S failed to send a message at 120 sec.
The figures below are the parts of the execution outputs.

• federate_s
  
• RTI
  

[edwardalee]

Interesting. This is probably some default timeout that needs to changed when opening a socket. I vaguely recall some sort of keepalive configuration option for sockets...

[byeonggiljun]

Thank you for getting interested in this, @edwardalee!

As you said, I could find a post that explains TCP keepalive (https://tldp.org/HOWTO/TCP-Keepalive-HOWTO/overview.html). I'll apply this option to the federated execution.

[edwardalee]

In addition to keepalive, we might consider a procedure that attempts to reconnect when it detects a failed connection. Of course, this won't work if the socket server isn't up, so I'm not sure what practical scenarios this will address.

[erlingrj]

I believe this is fixed with #422