About this article: This article is reviewed against the cited sources. This is source research, not lab validation. Check the stated versions and your own requirements before making changes.
Event 1311 (source NTDS KCC) is logged when Active Directory cannot work out a way to get every partition to every site that needs it. The message reads that "either (a) there is not enough physical connectivity published via the Active Directory Sites and Services Manager to create a spanning tree connecting all the sites containing the Partition…, or (b) replication cannot be performed with one or more critical servers" (Microsoft Learn, event 1311).
The Knowledge Consistency Checker (KCC) is the built-in process on every domain controller that builds replication connections. Between sites, one domain controller per site, the inter-site topology generator (ISTG), builds the connections using the sites, site links and costs you defined (Microsoft Learn, replication concepts). Event 1311 means that map does not join up, or a domain controller it depends on is not replicating.
What you see
- Event 1311 in the Directory Service log, usually on ISTG domain controllers, naming a partition (domain, schema, configuration, an application partition or the global catalog).
- Changes made in one site taking a long time, or never, to appear in another.
- Other replication errors on bridgehead domain controllers, which are often the real cause.
Why it happens
Microsoft groups the causes into two kinds (Microsoft Learn, event 1311).
The site design does not match the network:
- Site link bridging is on (the default) but the network is not fully routed, so the KCC builds a connection between two sites that cannot actually reach each other.
- One or more sites are not in any site link. This often happens when a site link is deleted and its sites are not added to another.
- Site links contain all the sites but do not join each other: the topology is in two separate pieces ("disjoint site links").
- An administrator set preferred bridgehead servers that do not hold the partition needed, or that are offline.
Infrastructure has failed:
- A WAN link is down.
- A domain controller holding a needed partition is offline.
- Replication is failing for one or more partitions, or a partner has outbound replication disabled.
- A hub bridgehead is overloaded, because it is undersized, has too many branch partners or the schedules are too aggressive.
- The KCC has built a path around a failed connection but keeps retrying the failed one every 15 minutes ("connection keeping" mode).
How to prove the cause
Microsoft's order of checks, each read-only until you change something:
1. Site-specific or forest-wide? Find the ISTG for each site (with Ldp, search CN=Sites,CN=Configuration,DC=<root> for (cn=NTDS Site Settings) and read interSiteTopologyGenerator). Check whether every ISTG logs 1311 or only some. A forest-wide pattern points at design; a single site points at that site's links or servers.
2. Read the site connectivity matrix. On the console of the ISTG (this command cannot be run remotely):
repadmin /showism
Each entry reads cost:interval:options. A site's entry for itself is 0:0:0. Positive numbers mean a working path. -1:0:0 means no path: the transport is not used, the site has no domain controllers ("uncovered", which is normal), or the site is not in a site link. With bridging on, a site that has domain controllers and shows -1:0:0 to every other site is orphaned. Put it in a site link.
3. Look for preferred bridgeheads.
ldifde -f SITEDUMP.LDF -d "CN=Sites,CN=Configuration,DC=<root domain>"
findstr /i "bridgeheadTransportList" SITEDUMP.LDF
Any server with bridgeheadTransportList set is a preferred bridgehead.
4. Find the replication failures behind it.
repadmin /failcache
repadmin /showrepl
dcdiag /test:intersite /e /q
dcdiag /test:connectivity /e /q
/failcache (run on each ISTG) lists the failures the KCC knows about; /showrepl shows which partition is failing. The two dcdiag tests check inter-site bridgeheads and name resolution and LDAP/RPC reachability to every domain controller, showing only those with errors.
5. Check for overloaded hubs. In repadmin /showrepl output, partners of an overloaded source show a last attempt of "never".
6. Draw the topology. For disjoint site links, Microsoft's advice is literally pencil and paper: list every site and site link and check that one connected graph covers them all. Do this last, after other causes are ruled out.
The fix
Work through these in the order the proof pointed to. After each change, Microsoft's guidance is to wait twice the longest replication interval in the forest before deciding whether 1311 has stopped.
- Bridging on a network that is not fully routed. Either make the network fully routed, or turn off Bridge all site links for the transport and create the site links and site link bridges that match real connectivity. Microsoft recommends keeping bridging on where the network is fully routed.
- Orphaned sites. Add each one to a site link with the site it is physically connected to.
- Preferred bridgeheads. Microsoft strongly recommends not defining them, because domain controllers already fail over between bridgeheads. In Sites and Services, open each domain controller's properties and remove the transports from "This server is a preferred bridgehead server for the following transports".
- Replication failures. Fix them, starting with domain controllers that hold a partition no one else holds. A domain controller that has been offline or failing for longer than the tombstone lifetime should not simply be brought back. Prove that first, read-only: check its last successful inbound replication with
repadmin /showrepl, and read the forest's tombstone lifetime withrepadmin /showattr <DC_name> "CN=Directory Service,CN=Windows NT,CN=Services,CN=Configuration,DC=<forest root domain>,DC=<top level domain>"(if the attribute is absent, the built-in default applies) (Microsoft Learn, error 8614). Microsoft adds: "Don't assume that a force demotion is the only workable solution." The clean-and-resume path is covered in the article on event 2042 and error 8614, listed with this one. Destructive step: removing the domain controller from the forest means a forced demotion, metadata cleanup and possibly seizing operations master roles it held. Any changes that never left it are lost. - Overloaded hub. Add capacity, spread branch sites across hub domain controllers or relax site link schedules.
- Disjoint site links. Add a site link that joins a site in each separate piece.
- Connection keeping mode. Delete the broken connection objects and let the KCC rebuild them.
How to confirm the fix
- No new 1311 events on any ISTG after twice the longest replication interval.
repadmin /showismon each ISTG shows positivecost:intervalvalues to every site that has domain controllers.repadmin /failcacheshows "No Failures" for the site's bridgeheads, anddcdiag /test:intersite /e /qreturns nothing.- A test change (for example a description edit on a test object) reaches each site within the expected schedule.
What to watch afterwards
- Any site or site link change. Deleting a site link is the classic way to orphan a site. Re-run
repadmin /showismafter every change to sites or links. - New and retired domain controllers. A site that loses its last domain controller becomes uncovered. Make sure every new site is placed in a site link, since Microsoft notes that sites not in a site link are left out of the topology.
- Hub load. As branch sites are added, check that hub bridgeheads still deliver to every partner.
Read next
Related articles you may want next.
- TroubleshootingAD replication error 1722 "The RPC server is unavailable": find the layer that is really failingActive Directory replication fails with 1722 (0x6ba). Prove whether name resolution, a blocked port, signing or the network is at fault, fix that layer and confirm replication has caught up.Read article
- TroubleshootingEvent 2042 / error 8614 "It has been too long since this machine replicated": rejoin or remove the domain controllerA domain controller stopped replicating with event 2042 or error 8614 because a partner was silent longer than the tombstone lifetime. Decide between removing the domain controller and cleaning lingering objects, then restore protection.Read article
- HostingOne server per role or combined: domain controller, file, application and SQLHow a small business decides which server roles can share a hosted virtual machine and which must be kept apart, with Microsoft's explicit cautions about domain controllers and SQL Server.Read article
Risks and limits
- Do not turn off site link bridging as a quick fix in a fully routed network. You then have to define every transitive path by hand, and missing one causes the same event.
- Delete only broken connection objects. The KCC does not change manual or administratively modified connection objects (Microsoft Learn, replication concepts), so a broken manual connection has to be dealt with by you.
- Microsoft notes that SMTP replication will not be supported in future versions of AD DS, so do not create new SMTP site links (Microsoft Learn, replication concepts).
- The troubleshooting page quotes a default tombstone lifetime of 60 days; forests created on currently supported versions default to 180 days. Read your forest's actual value before deciding a domain controller is beyond recovery.
Sources
- Microsoft Learn, "How to troubleshoot Event ID 1311 messages on a Windows domain" (KB 307593), last updated 12 February 2026: https://learn.microsoft.com/en-us/troubleshoot/windows-server/active-directory/troubleshoot-event-id-1311-messages
- Microsoft Learn, "Active Directory Replication Concepts", last updated 12 May 2025: https://learn.microsoft.com/en-us/windows-server/identity/ad-ds/get-started/replication/active-directory-replication-concepts
- Microsoft Learn, "How to detect and remove lingering objects in a Windows Server Active Directory forest" (tombstone lifetime default), last updated 12 February 2026: https://learn.microsoft.com/en-us/troubleshoot/windows-server/active-directory/information-lingering-objects
- Microsoft Learn, "Ldifde" (command syntax for
-fand-d), Windows Server 2012 R2 documentation, last updated 31 August 2016: https://learn.microsoft.com/en-us/previous-versions/windows/it-pro/windows-server-2012-r2-and-2012/cc731033(v=ws.11) - Microsoft Learn, "Active Directory replication error 8614" (repadmin /showattr for tombstone lifetime; forced demotion): https://learn.microsoft.com/en-us/troubleshoot/windows-server/active-directory/replication-error-8614
Next step
Event 1311 is often the first sign that a hosted or branch site was added without its site links being thought through. If you want your sites and links checked against the real network and replication monitored afterwards, NetLeverage has looked after business servers since 2002. Talk to us through managed support with the repadmin /showism output from an affected ISTG.
This article was originally published on .