Subnets are representing associated IP address ranges in each site. This is equal to subnets allocated in network devices, but does not need to be exactly the same. For example, if a site uses 10-20 class C subnets, then instead of adding all these to Active Directory sites subnet, we can summarize all those into a class B subnet and use them. Based on this subnet information, it allows objects to locate the closest domain controller. When physical subnets are added or removed, they should be updated in the Active Directory site configuration as well. Otherwise, they will pass unwanted traffic via slow links.