Skip to content

How to Configure Active Directory Sites and Subnets Print

  • 0

Active Directory sites describe the network locations that clients and domain controllers use to choose nearby services and plan replication. A site is not simply an office name: its subnet objects must accurately represent routed client networks.

Map the network before editing the directory

Collect each location’s routed IP prefixes, the domain controllers serving it, WAN connectivity and the intended replication schedule. Confirm whether VPN users enter through a central network or retain location-specific address ranges. Missing or overlapping subnet definitions can send clients to a distant controller even when a nearby controller is healthy.

Use the ActiveDirectory module with an account delegated to manage sites and subnets. Coordinate with the network team so the prefix reflects the actual mask used by clients. A label such as Toronto is only an example and does not establish where a hosted service is physically located.

Inspect the existing topology

Review sites and their subnet mappings before creating new objects. Also open Active Directory Sites and Services to inspect the existing server objects and site links. Record the current configuration so changes can be reversed deliberately.

Get-ADReplicationSite -Filter * | Select-Object Name
Get-ADReplicationSubnet -Filter * -Properties Site | Select-Object Name,Site
Get-ADReplicationSiteLink -Filter * -Properties Cost,ReplicationFrequencyInMinutes

Create a site and associate a subnet

The following example creates a site and maps a documentation-only network prefix to it. Substitute the approved site and network. Creating these objects does not move a domain controller or create the required network connectivity.

New-ADReplicationSite -Name 'Toronto-Office'
New-ADReplicationSubnet -Name '192.0.2.0/24' -Site 'Toronto-Office' -Location 'Toronto office'

Place controllers and review site links

In Active Directory Sites and Services, move each affected server object to the site that corresponds to its actual network location. Review the intersite transport links, their member sites, costs and schedules. A cost expresses route preference for replication; it does not configure an IP route or reserve bandwidth. Avoid changing every schedule merely to make replication appear immediate.

Apply a small, documented change first. Let it replicate and check the result from another controller before adding the remaining subnets. Keep a controller available to serve clients while topology changes settle.

Verify from a representative client

Run the site-discovery command on a client with an address in the mapped subnet. Compare the returned site with the expected one, then confirm the domain controller selected for the domain. Recheck replication health after moving controller objects.

nltest /dsgetsite
nltest /dsgetdc:corp.example
repadmin /replsummary

Troubleshooting and rollback

If a client reports the wrong site, verify its active address and subnet mask, VPN interfaces and the prefix stored in Active Directory. Do not infer site selection solely from the building where the user works. If authentication fails after moving a controller, investigate DNS and connectivity rather than repeatedly relocating the server object.

Rollback means restoring the previous subnet association, server placement and site-link settings. Delete a newly created site only after confirming that it has no dependent server or subnet objects. Keep the original mapping and the verification output in the change record.

Microsoft reference

Use the Microsoft ActiveDirectory PowerShell reference to confirm the parameters for New-ADReplicationSite and New-ADReplicationSubnet on the management host.


Was this answer helpful?

« Back