AD Replication Error 8614 and Event 2042: Tombstone Lifetime Exceeded

Fix Active Directory replication error 8614 and NTDS Replication event 2042 (time since last replication exceeded the tombstone lifetime): find the cause, check for time jumps, remove lingering objects, then lift the quarantine safely.

Home › Guides › AD replication error 8614

Last updated 18 November 2026

Quick answer

Error 8614 (0x21a6) and NTDS Replication event 2042 mean a destination DC has stopped replicating from a source DC because it hasn't inbound-replicated that directory partition for tombstone lifetime (TSL) days, or because its clock jumped so far that it looks that way. The block is deliberate: it stops lingering objects (deleted objects that still exist on the out-of-date DC) from spreading.

Microsoft gives two ways out:

Don't set the registry value first. Microsoft's order is lingering objects out, then lift the quarantine.

What the errors look like

dcdiag (Replications test):

The replication generated an error (8614):
Active Directory cannot replicate with this server because the time since the last replication with this server has exceeded the tombstone lifetime.

repadmin /showreps (also seen with /replsum, /showrepl and /syncall):

Last attempt @ <date> <time> failed, result 8614(0x21a6):
The Active Directory cannot replicate with this server because the time since the last replication with this server has exceeded the tombstone lifetime.

Replicate now in Active Directory Sites and Services fails with the same sentence. In the Directory Service log, NTDS Replication event 2042:

It has been too long since this machine last replicated with the named source
machine. The time between replications with this source has exceeded the tombstone
lifetime. Replication has been stopped with this source.
The reason that replication is not allowed to continue is that the two machine's views of deleted objects may now be different.

The event also lists the time of the last successful replication, the source's invocation ID and name, and the tombstone lifetime in days. The article also lists NTDS KCC event 1925: "The attempt to establish a replication link for the following writable directory partition failed."

Why it happens

When an object is deleted, AD DS keeps a tombstone that other DCs replicate in; it stays for the tombstone lifetime and is then garbage-collected. A DC that misses replication for longer than that never hears about some deletions, so it still holds objects that are gone everywhere else: lingering objects. Microsoft lists two root causes for 8614 and 2042:

Microsoft also notes that 8614 on a destination DC doesn't mean the fault is on that DC: it can be the network, DNS name resolution, authentication, the Jet database, topology or the replication engine on either DC. While 8614 is logged, earlier replication errors from the previous TSL days are masked.

Step 1: Check the tombstone lifetime

repadmin /showattr <DC_name> "CN=Directory Service,CN=Windows NT,CN=Services,CN=Configuration,DC=<forest root domain>,DC=<top level domain>"

Look for tombstoneLifetime. If it isn't in the output, an internal default is used. The 8614 article says the default is 60 or 180 days depending on the Windows version; the lingering objects article says it depends on the first DC installed in the forest, is 180 days for all currently supported Windows Server versions, and doesn't change when you upgrade DCs. A short custom value (Microsoft's example is two days) makes 8614 much easier to hit.

Step 2: Find every DC that failed for TSL days

repadmin /showrepl * /csv

Open the output in Excel and sort on the last success column, oldest first. Grouping the failing DCs by site can show a WAN link or site problem; map their addresses to site subnets with the IPv4 Subnet Calculator or IPv6 Subnet Calculator. Events that warn of the same problem before 8614 (from the lingering objects article):

EventMeaning
1862, 1863The local DC has not recently received replication information from several DCs (inter-site)
1864The local DC has not recently received replication information from several DCs (summary)
1311The KCC was not able to build a spanning tree topology
2042It has been too long since this server last replicated with the named source server

Step 3: Look for time jumps

On the DCs logging 8614, check event logs and repadmin/dcdiag output for date stamps from before the OS was released or installed in your forest, date stamps in the future, and date ranges with no events at all. Microsoft Support has seen DC clocks jump "hours, days, weeks, years, and even tens of years in the past and future". If the time is wrong, fix it, then find out why. Microsoft's checklist:

The w32tm checks and settings are in Kerberos clock skew (KRB_AP_ERR_SKEW).

Step 4: Remove lingering objects

Microsoft's preferred tool is Lingering Object Liquidator v2 (LoLv2). If you can't use it, use repadmin:

repadmin /removelingeringobjects <Dest_DSA_LIST> <Source DSA GUID> <NC> [/advisory_mode]
repadmin /help:removelingeringobject

Signs that lingering objects are already present (lingering objects article): events 1084 (There is no such object on the server), 1388 (This destination system received an update for an object that should have been present locally but was not) and 1311, and repadmin errors 8240 (There is no such object on the server) and 8606 (Insufficient attributes were given to create an object). A DC that logs event 1988 doesn't hold the lingering object; its source DC does. Other symptoms include deleted users or groups still in the Exchange GAL, duplicate objects, and errors that an object already exists when you can't see it.

Step 5: Check strict replication consistency

With Strict Replication Consistency enabled, a destination DC that receives an update for an object it doesn't have stops inbound replication of that partition from the source. With it disabled, the destination requests the full object and a deleted object reappears in the forest. It defaults to 1 on forests created by promoting Windows Server 2003 or later, and 0 if the forest was created on Windows 2000 Server; raising the functional level doesn't change it.

Path: HKEY_LOCAL_MACHINE\system\ccs\services\ntds\parameters
Setting: Strict Replication Consistency
Type: reg_dword
Value: 0 | 1
repadmin /regkey <fully qualified computer name> +strict
repadmin /regkey * +strict
repadmin /regkey gc: +strict

(One DC, all DCs in the forest, all GCs in the forest.)

Step 6: Lift the quarantine, fix replication, put it back

  1. Only after lingering objects are removed, set Allow Replication With Divergent and Corrupt Partner to 1 on the DC logging 8614:
    Registry path: HKEY_LOCAL_MACHINE\system\ccs\services\ntds\parameters
    Registry setting: Allow replication with divergent and corrupt partner
    Registry value: 0 = disallow, 1 = allow
    or with repadmin:
    repadmin /regkey dc01.contoso.com +allowDivergent
    repadmin /regkey * +allowDivergent
    repadmin /regkey GC: +allowDivergent
  2. Resolve the replication failures that are now visible. Review past Directory Service events and dcdiag and repadmin output from the source DC, the destination DC and other partners.
  3. Delete the setting or set it back to 0. Event 2042's own guidance is to remove the key once the systems have replicated once, to reinstate the protection.

Step 7: Stop it happening again

Use the PowerShell Cheat Sheet for Get-ADDomainController and Get-WinEvent one-liners (for example listing every DC to check), filter exported repadmin CSV or event text with the Regex Tester, and look up 2042, 1388 and 1925 in the Windows Event ID Lookup.

Related tools and guides

Look up 2042, 1925, 1388 and 1311 in the Windows Event ID Lookup, list your DCs with Get-ADDomainController from the PowerShell Cheat Sheet and filter repadmin /showrepl CSV output with the Regex Tester. Related guides: Kerberos clock skew, DFSR SYSVOL events 2213 and 4012 and the trust relationship between this workstation and the primary domain failed.

FAQ

What does Active Directory replication error 8614 mean?

The destination DC refuses to replicate from a source because the time since their last successful replication exceeds the tombstone lifetime. The two DCs may now disagree about deleted objects, so replication is quarantined to stop lingering objects spreading. NTDS Replication event 2042 reports the same condition.

What causes error 8614 and event 2042?

Microsoft lists two root causes: the destination DC failed to inbound-replicate a partition from a source for tombstone lifetime days, or its system time jumped tombstone lifetime days or more since the last successful replication. The underlying fault can be on the source DC, the destination DC or the network.

What is the default tombstone lifetime?

Microsoft's lingering objects article says it depends on the first DC installed in the forest and is 180 days for all currently supported Windows Server versions; the 8614 article says 60 or 180 days depending on version. An existing value doesn't change when you upgrade DCs. Check it with repadmin /showattr on CN=Directory Service.

Can I just set Allow Replication With Divergent and Corrupt Partner to 1?

Not as a first step. Microsoft's plan is to check for and remove lingering objects first, then set the value to 1 on the DC logging 8614, fix replication, and delete the value or set it back to 0. Otherwise deleted objects can reappear in the forest.

Should I force-demote a DC with error 8614?

It is one of Microsoft's two options. It's easier and faster than recovering in place, but originating updates that exist only on that DC can be lost. Microsoft suggests considering it for DCs that reach 90 percent of the tombstone lifetime without replicating.

Want the full kit?

Runbooks, checklists and helpdesk templates: IT Admin Packs on Gumroad.