Domain Controller Replication Troubleshooting

Domain Controller Replication Troubleshooting

1. - Check DFS Replication on all domain controllers

Check if domain controllers are sharing SYSVOL, have they recently logged an Error event, and are in an "error" state;

Check for the SYSVOL share

You may manually check whether SYSVOL is shared or you can run the following command to inspect each domain controller by using the net view command:

For /f %i IN ('dsquery server -o rdn') do @echo %i && @(net view \\%i | find "SYSVOL") & echo 

Check DFS Replication state

To check DFS Replication's state on domain controllers, you may query WMI. You can query all domain controllers in the domain for the SYSVOL Share replicated folder by using WMI as follows:

For /f %i IN ('dsquery server -o rdn') do @echo %i && @wmic /node:"%i" /namespace:\\root\microsoftdfs path dfsrreplicatedfolderinfo WHERE replicatedfoldername='SYSVOL share' get replicationgroupname,replicatedfoldername,state 

The "state" values can be any of the following:

0 = Uninitialized

1 = Initialized

2 = Initial Sync

3 = Auto Recovery

4 = Normal

5 = In Error

Note Depending on a domain controller's condition, it may fail to report a state value and indicate "No Instance(s) Available."

Check Event logs for recent errors or warnings

If any domain controllers do not report the "SYSVOL Share" replicated folder as being in a state "4" (normal), check the event log of those domain controller(s) to evaluate their condition. Review each domain controller for recent errors or warnings in the DFS Replication event log, such as the warning event ID 2213, which indicates that DFS Replication is currently paused.

Check the Content Freshness configuration

Determine whether DFS Replication triggered content freshness protection on the affected domain controllers. Content Freshness is enabled on Windows Server 2012 (and later versions) domain controllers by default, but may also be manually enabled on Windows Server 2008 and 2008 R2 servers.  

To evaluate if content freshness is enabled, the MaxOfflineTimeInDays setting will be set to 60. If content freshness is disabled, MaxOfflineTimeInDays will be set to 0. To check MaxOfflineTimeInDays, run the following command:

wmic.exe /node:%computername% /namespace:\\root\microsoftdfs path DfsrMachineConfig get MaxOfflineTimeInDays 

To query all domain controllers in the domain, run the following command:

For /f %i IN ('dsquery server -o rdn') do @echo %i && @wmic /node:"%i" /namespace:\\root\microsoftdfs path DfsrMachineConfig get MaxOfflineTimeInDays 

For each domain controller enabled for content freshness, evaluate if DFS Replication has logged an event ID 4012 that indicates replication of the folder has stopped because replication has failed for longer than the MaxOfflineTimeInDays parameter.  

2. Prepare the domain controllers that are in an error state

Install appropriate updates

For any domain controllers running Windows Server 2008 or 2008 R2, first install DFS Replication updates to prevent data loss and to fix known issues. It is a best practice to use the latest version of DFS Replication. The latest version of DFS Replication is discussed in article 968429. 

Back up SYSVOL data

Perform a backup of SYSVOL data (if present) on each domain controller. Backups may be a simple file copy of the SYSVOL contents to a safe location or, it may be a backup that uses backup software.  

Depending on the situation, policy files could be moved to "PreExisting" or "Conflict and Deleted." "PreExisting" and "Conflict and Deleted" contents will be purged if initial synchronization is performed multiple times on a server. Back up data in these locations to avoid data loss.

3. Recover DFS Replication on the domain controllers in the error state

Check if a dirty shutdown was detected and whether DFS Replication is paused on any domain controllers (event ID 2213). You may find a domain controller is waiting to complete initialization of SYSVOL after promotion. It will have logged a 4614 event that indicates that DFS Replication is waiting to perform initial replication, and it will not have logged a 4604 event signaling that DFS Replication has initialized SYSVOL.

If content freshness is enabled and there are three or more domain controllers in the domain

Content freshness protection will log an event ID 4012 that indicates that replication has stopped because replication on the folder has failed for longer than the MaxOfflineTimeInDays parameter. You must follow the instructions in article 2218556to reinitialize DFS Replication on the affected domain controller(s).  

If all domain controllers have logged the 4012 event and their "state" is "5," then follow the instructions in article 2218556to completely initialize SYSVOL. This is the only situation to set a DFS Replication server as authoritative. Make sure that the domain controller configured as authoritative has the most up-to-date copy of all SYSVOL contents.

Or, if one or more domain controllers are blocking replication because of content freshness, they each must be non-authoritatively recovered. To do this, follow these steps:

1.     Backup all SYSVOL contents of the domain controller(s). Typically, policy edits are performed on the PDC Emulator, but this is not guaranteed. Any data present on the recovered domain controller(s) not matching the partners will go into the "PreExisting" or "Conflict and Deleted" folder, or both.

2.     Next, set the domain controller(s) as nonauthoritative by disabling the membership as described in article 2218556. You must be aware of the replication topology, and you must "fan out" from a healthy domain controller by selecting direct partners of it, then recovering further downstream domain controllers, and so on. Event ID 4144 will be logged to confirm the membership is disabled. Make sure all domain controllers requiring recovery log this event. It may be necessary to force Active Directory replication and then run the dfsrdiag pollad command on each domain controller to detect the disabled membership quickly.

3.     Enable the membership and wait for the 4614 and 4604 events to report completion of the initial synchronization. Restore any required files from backup or from "PreExisting" and "Conflict and Deleted" as necessary.

If content freshness is not enabled or triggered and there are three or more domain controllers in the domain

If content freshness protection was not triggered, run the ResumeReplication WMI method on the affected domain controllers. You must be aware of the replication topology, and you must "fan out" from a healthy domain controller by selecting direct partners of it, then recovering further downstream domain controllers, and so on. After replication is resumed, DFS Replication will log events 2212, 2218, and then 2214 (indicating that DFS Replication initialized the SYSVOL replicated folder).  

Preventing future occurrences of the issue

Check whether the Application and System event logs are frequently reporting ESENT database recovery operations, disk performance problems, or both. These typically coincide with unexpected shutdowns of the system, with DFS Replication not stopping gracefully, or disk subsystem failures. Consider updating the system's drivers, installing appropriate updates to the disk subsystem, or contacting the system's hardware manufacturer to investigate further. You may also contact Microsoft Customer Support Services to help evaluate the system's health and DFS Replication behavior.

The Service Control Manager (SCM) uses the default time-out time of 20 seconds for stopping a service. In some complex DFS Replication implementations, this time-out value may be too short, and DFS Replication stops before the appropriate database is closed. At service restart, DFS Replication detects this condition, and then performs the database recovery. WaitToKillServiceTimeout may be used to grant DFS Replication more time to commit changes to the database during shutdown. For more information, go to article 977518.

After you have restored DFS Replication of SYSVOL, DFS Replication health must be carefully monitored in the environment to prevent this scenario. Regular review of DFS Replication event logs, collecting of DFS Replication health reports, and collecting of replication state (by using the WMI query in the "Check DFS Replication state" section under "Step 1 - Evaluate the state of DFS Replication on all domain controllers" in this article) are recommended.