I have just completed the process of migrating from Exchange 2007 Standard SP3 RP10 to Exchange 2013 Standard. I now have Exchange 2013 Standard CU3 installed on two Server 2012 Standard systems. The two new exchange servers are members of a DAG.
With Exchange 2007, I had to include the IP addresses of multi-function devices (MFDs) with scan to e-mail capability in a receive connector in order for the scan to e-mail to work. Five of the MFDs were on the same subnet as our Exchange 2007 servers, and one MFD was not. It is on the same subnet as our Exchange 2013 servers.
I created new Receive connectors on our Exchange 2013 servers for this purpose. The one MFD on the same subnet as the Exchange 2013 servers was mistakenly excluded from a receive connector. Much to my surprise, it is able to relay mail.
I removed the IP address of one of the other MFDs to see if the fact that it is on a different subnet than the Exchange 2013 servers made a difference. I replicated our domain controllers to help ensure that a lack of AD replication would not be an issue with this test. The excluded MFD is still able to relay items.
I have an application on a server. The application relays messages, but the IP address of the server it is on is not included in any receive connectors.
By searching through the SMTPreceive log, I found the IP address of the MFDs I did my scan to e-mail tests from in the log. I also found the IP address and relayed messages from the application server. The log indicates that the Default Frontend connector is handling this traffic. Does Exchange 2013, by default, relay messages from all internal IP addresses? Does this mean that I don’t need receive connectors for any internal devices (MFDs, servers, etc.). Our Exchange 2007 servers have been decommissioned. Would I have needed connectors while we were still in a co-existence situation? Am I likely to need them when we have our next co-existence situation in a few years?