Sometimes error messages can lead you down the wrong path… sometimes messages, errors or others, aren’t captured and returned in whole to the application layer. Here is an example. This error was received by a vendor application which was attempting to connect to an Oracle database:
Connection to <hostname>:80 closed after being inactive for 129 seconds, exceeding the “HTTP Keep-Alive timeout”
This may look like a situation that requires re-configuration of firewall or SqlNet timeout parameters, or the implementation of a keep-alive process. But, in this case the connection timed out because it was waiting for a response to the prompt for a new user password.
So, before contacting the network and firewall support people, contact your DBA and have them check to see if your userid is in it’s expired grace period.