AWS session time out FIX

Assuming your Amazon EC2 instance is running Linux (and the very likely case that you are using SSH-2, not 1), the following should work pretty handily:

1. Remote into your EC2 instance.

   ssh -i <YOUR_PRIVATE_KEY_FILE>.pem <INTERNET_ADDRESS_OF_YOUR_INSTANCE>
   

2. Add a "client-alive" directive to the instance's SSH-server configuration file.

   echo 'ClientAliveInterval 60' | sudo tee --append /etc/ssh/sshd_config
   

3. Restart or reload the SSH server, for it to recognize the configuration change.
   • The command for that on Ubuntu Linux would be..

     sudo service ssh restart
     

   • On any other Linux, though, the following is probably correct..

     sudo service sshd restart
     

4. Disconnect.

   logout
   

The next time you SSH into that EC2 instance, those super-annoying frequent connection freezes/timeouts/drops should hopefully be gone.

This is also helps with Google Compute Engine instances, which come with similarly annoying default settings.

Warning: Do note that TCPKeepAlive settings (which also exist) are subtly, yet distinctly different from the ClientAlive settings that I propose above, and that changing TCPKeepAlive settings from the default may actually hurt your situation rather than help.

More info here: http://man.openbsd.org/?query=sshd_config