Home Runbooks AD Replication Errors: What the Event IDs Actually Mean
Troubleshooting FREE

AD Replication Errors: What the Event IDs Actually Mean

AD replication errors fill the event log with cryptic codes. This runbook decodes the most common ones and gives you the exact fix for each.

⌛ 10 min read· Updated 2026

Run Repadmin First

Before looking at individual event IDs, get the full picture:

# Overall replication summary — shows failures at a glance
repadmin /replsummary

# Show all replication partners and last successful replication time
repadmin /showrepl

# Show replication queue (backlog of changes waiting)
repadmin /queue

# Export failures to CSV for review
repadmin /showrepl * /csv > repl-report.csv

The replsummary output tells you which DC is failing, which partner, how long it's been failing, and the error code. Start there before reading event logs.

Event 1864 — Replication Latency

Event 1864 fires when a DC has not replicated in 8+ days. After 60 days, tombstone lifetime can expire and manual intervention is required.

# Check when a DC last replicated
repadmin /showrepl DC01 /verbose

# Check tombstone lifetime for your forest
(Get-ADObject -Identity (Get-ADRootDSE).configurationNamingContext `
  -Properties tombstoneLifetime).tombstoneLifetime
WarningA DC that hasn't replicated past the tombstone lifetime (usually 180 days) must be forcibly demoted, not fixed. Bringing it back into replication will cause deleted objects to resurrect across the forest.
# After resolving the underlying error, force sync
repadmin /syncall /AdeP

Error 8606 — Lingering Objects

An object was deleted on one DC but another DC that was offline missed the deletion and still holds a copy. When replication resumes, the DCs disagree about whether the object exists.

# Detect lingering objects — advisory mode first (no changes)
repadmin /removelingeringobjects DC02 DC01 "DC=corp,DC=local" /advisory_mode

Review the output, then remove for real:

repadmin /removelingeringobjects DC02 DC01 "DC=corp,DC=local"
repadmin /removelingeringobjects DC02 DC01 "CN=Configuration,DC=corp,DC=local"
repadmin /removelingeringobjects DC02 DC01 "CN=Schema,CN=Configuration,DC=corp,DC=local"

Error 1722 — RPC Server Unavailable

RPC 1722 means the source DC is unreachable or RPC ports are blocked. This is almost always a firewall or network issue.

# Test RPC connectivity
Test-NetConnection -ComputerName DC01 -Port 135
Test-NetConnection -ComputerName DC01 -Port 49152  # Start of dynamic port range

AD replication uses port 135 plus dynamic ports 49152–65535. To pin replication to a specific port:

# Set a static RPC port for AD replication
reg add HKLM\SYSTEM\CurrentControlSet\Services\NTDS\Parameters /v "TCP/IP Port" /t REG_DWORD /d 50000 /f
net stop netlogon && net start netlogon

Error 8524 — DNS Lookup Failed

The source DC can't be resolved by name. Replication requires DCs to find each other via DNS SRV records.

# Verify the source DC resolves
Resolve-DnsName DC01.corp.local

# Check _msdcs zone exists
Get-DnsServerZone | Where-Object {$_.ZoneName -like "*msdcs*"}

# Re-register DC DNS records
ipconfig /registerdns
net stop netlogon && net start netlogon

Error 1753 — Endpoint Mapper Failed

RPC port 135 is reachable but the service being called is not registered. Usually caused by AD DS not fully starting.

# Check if AD services are healthy on the source DC
Get-Service NTDS, ADWS, Netlogon, DNS, KDC | Select Name, Status

# Restart the AD DS service if it's stopped or degraded
net stop NTDS && net start NTDS

Force Replication After Fixing

# Force replication of a specific partition
repadmin /replicate DC02 DC01 "DC=corp,DC=local"

# Force sync of all partitions across all DCs
repadmin /syncall /AdeP

# Verify it worked
repadmin /showrepl | findstr /i "error|fail|last attempt"
TipRun repadmin /replsummary again after forcing. The failure count and last error time should reset. If they don't, the underlying issue isn't fully resolved.