I understand that for classful networks in EIGRP we do not need a subnet mask. Why is this?
When I do a show ip protocols
the routes appear without a subnet mask which looks somewhat odd.
If I have a 192.168.0.0/16 network and just introduce the network into the EIGRP routing protocol as 192.168.0.0, how will EIGRP know that it is 192.168.0.0/16 and not say 192.168.0.0/24?
Or does EIGRP understand the network address from the host addresses assigned to the router's DC interfaces?
Thanks.
Best Answer
By default EIGRP uses a classful network, meaning that it assumes that an address in the class A range with have an 8-bit subnet mask. There is no automatic process of checking other than the assumption that if you do not specify the subnet mask in wildcard mask format ( host mask ) then you want to use the default subnet mask for the IP range you've specified.
http://study-ccna.com/eigrp-configuration
As @Todd Wilcox pointed out below, each address range has a class, originally specified in RFC791 ( https://www.ietf.org/rfc/rfc791.txt ), excluding special use addresses ( https://www.rfc-editor.org/rfc/rfc3330 ).
Basically:
So converting the first octet ( first group of decimal digits in an IP address: 192.168.1.0 ) into binary gives us 11000000, matching the MSB ( most significant bits ) of a C class address, implying a default subnet mask of /24 bits or 255.255.255.0.