Many people want to use multiple IP addresses on a single NIC interface with Exchange 2013 and Windows Server 2012. There are several reasons for this multi-homed IP configuration, such as various receive connectors for diverse applications (fax, SharePoint, gateways, etc.), or for an additional IIS website, amongst other things.
The behavior that changed since Windows Server 2008 is that the source IP address on a NIC will always be the lowest numerical IP address. The challenge for every Exchange administrator is to configure the correct firewall settings for the appropriate IP address. Lets make an example:
You install the first Exchange 2013 server in your environment with the „primary“ IP address of 10.35.3.200. Afterwards you decide to add an additional IP address to your MAPI network adapter, such as for a further receive connector. The secondary IP address would be 10.35.3.100. No you have two issues:
- The server registered two DNS names for the same server
- The server is now sending all outbound traffic from 10.35.3.100, because 100 is lower than 200.
Let’s go a bit deeper and explain the “weak host model“ and the “strong host model“ for multi-homed servers and how they choose the source IP address selection.