Skip to content
INSIGHTS & GUIDES · NETCLOUD24

Reading Repadmin evidence before repairing AD replication

Diagnose replication from evidence about partners, naming contexts and timing. Use Repadmin read-only queries, correlate DNS, network and authentication findings, and verify the actual correction before broader repair.

Read the article ↓
Editorial illustration: How to Diagnose Active Directory Replication with Repadmin

A replication warning identifies a relationship that needs investigation. It does not establish that forcing synchronisation, restarting services or rebuilding a controller is the right remedy. Use Repadmin to locate the failing destination, source, naming context and timeline before choosing a repair.

This diagnostic guide applies to authorised AD DS administrators supporting Windows Server 2019, 2022 or 2025. Netcloud24 handles agreed directory administration in Managed services. Customers should report affected sign-ins or stale information rather than run replication repair commands independently.

Define the observed impact

Record which users, sites or applications are affected and when the issue began. A password change working at one office but not another can suggest a different path from a forest-wide controller outage.

List recent network, DNS, controller, certificate or security-policy changes. Report the sequence without assuming the latest event is necessarily the cause.

Preserve initial diagnostics before changes. A forced operation can replace useful failure timing with a newer result while leaving the underlying fault unchanged.

Use an authorised diagnostic context

Repadmin is available with the relevant AD administration tools. Use a supported controller or RSAT management device and the rights needed to inspect the intended directory scope.

Confirm domain and forest context, DNS and time. Avoid collecting a huge cross-forest report when the incident concerns one known pair, especially where reports expose internal topology.

Keep controller names, collection time and command with each output. A screenshot of a numeric error without its source and destination can be difficult to interpret later.

Begin with the replication summary

repadmin /replsummary

Review reported failures, participating controllers and the largest deltas. A delta reflects time since relevant successful replication, not automatically a universal failure threshold.

Interpret it against topology, site-link schedules and the expected change flow. A quiet scheduled intersite connection can have different timing from an always-available intrasite path.

Identify whether failures cluster around one source, one destination or a network boundary. That pattern helps focus the next query and prevents treating every listed controller as independently broken.

Inspect the failing controller and partner

repadmin /showrepl UK-DC02
repadmin /showrepl UK-DC01

The names are illustrative. Use the actual healthy and failing participants identified in the summary. Read inbound partner entries, last success, last attempt and the complete status text.

Replication is directional. A healthy inbound connection on one controller does not prove the reverse or every other naming context is healthy. Compare the exact source and destination relationship.

Distinguish domain, configuration, schema and DNS application partitions. A failure limited to one naming context can require a different investigation from a total network or authentication loss.

Separate the diagnostic categories

Name resolution: verify relevant controller and directory locator records through approved resolvers. A website resolving successfully does not prove AD replication can locate its partner.

Network and RPC: inspect routing, firewall changes and required directory communication. An ICMP response alone does not validate the full replication path.

Authentication and permissions: check time accuracy, trust context and relevant security changes. Do not apply broad permission grants or disable authentication to suppress an error.

Directory state: investigate database, topology or object-consistency errors through their specific supported procedures. A persistent naming-context problem may need specialist directory recovery, not a service restart.

Correlate events and other targeted checks

Review Directory Service, DNS Server and relevant system events around the failed attempt. Preserve messages and identifiers rather than only the event number.

dcdiag /test:dns /v
w32tm /query /status

These checks support DNS and time investigation but do not prove every replication component is healthy. Run them in the intended context and interpret their output alongside the failing relationship.

Check platform health if a controller was paused, disconnected, restored or moved. Report those operations to the directory owner because they can matter to supported recovery decisions.

Use change evidence rather than a generic reset

If a firewall rule changed immediately before failure, verify the actual required path and restore the approved rule through the network process. Do not open all traffic between all servers as a permanent solution.

If DNS records are wrong, correct their authoritative source and registration owner. Repeated client cache clearing does not repair an invalid directory record.

If a controller has been offline too long or reports object-consistency safeguards, stop generic synchronisation attempts and follow the documented error-specific recovery. Do not disable protections to force stale state back into the directory.

Force replication only for a defined reason

Repadmin includes operations that request synchronisation, but those are not first-line repairs for every red summary entry. They do not fix routing, authentication or corrupted state.

After the root cause is corrected, an authorised targeted request may help verify the intended relationship without waiting for the normal schedule. Select its source, destination and naming context deliberately and understand the operation before executing it.

A broad copied syncall command can affect a much wider scope than the incident requires. Keep routine diagnostics read-only until the repair and verification plan are approved.

Verify recovery over the expected interval

Repeat the same focused summary and partner queries after the correction. Compare timestamps and failures with the baseline, allowing the topology’s expected schedule.

Verify a representative safe directory change or application behaviour through the organisation’s supported test process. Do not modify production passwords, memberships or schema merely to create a diagnostic event.

Confirm monitoring no longer reports persistent failures and retain evidence from relevant partners. A single successful attempt should not hide another naming context or site still failing.

Escalate with a useful incident record

Provide domain and controller context, failure window, partner direction, naming context, complete error and recent changes. Protect topology and account data when transferring logs.

For forest-wide failure or suspected compromise, involve the directory recovery and incident owners. Routine replication repair is not a substitute for assessing an untrusted directory.

See controller health prerequisites and Microsoft’s replication failure diagnosis and replication troubleshooting guidance.


Was this answer helpful?

« Back
Contact

Write to us

Do you have questions about our services or need a quote? Write to us and we will respond as soon as possible.

Contact form

Fields marked * are required.