Any DFS gurus out there?

[Guldan]

TL;DR my datacenter lost power, Yay! Most things came up OK but DFS-R was borked and I didn't get an alert because my email server was down as well. I have two identical Dell R510 servers with about 3-4TB total on them.

So replication stopped for a month and one of my servers is way behind now, Luck for me I was smart enough not to do active/active and all the LIVE data was to one server (not both) so I don't have a split brain scenario.

On the server that is way behind I killed all the data, I've delete all the replication groups etc,. I tried to create new replication groups but it won't replicate because I have to do that WMI stuff to start it again.. But I don't want to re-start replication I want to re-make it completely if that makes sense.

How do I 100% kill and remake DFS-R cleanly? Should I remove-re-install the role? kill the registry key? I just want to re-sync everything from scratch.

 

[Zedicus]

on the out of sync box i would go through the roles manager and actually remove DFS and all the features it adds. then reboot, and reinstall. that should be enough to get it to think it will be a new DFS replicate. (on the main box make SURE you removed the DFS sync partner before uninstalling DFS on the backup)

 

[Guldan]

on the out of sync box i would go through the roles manager and actually remove DFS and all the features it adds. then reboot, and reinstall. that should be enough to get it to think it will be a new DFS replicate. (on the main box make SURE you removed the DFS sync partner before uninstalling DFS on the backup)

I killed all DFSR groups and even deleted all the DFSRprivate folders and data, users have no idea what happened (which is great)

Good call on removing it on the dead server, if that works you're a hero.

 

[Guldan]

on the out of sync box i would go through the roles manager and actually remove DFS and all the features it adds. then reboot, and reinstall. that should be enough to get it to think it will be a new DFS replicate. (on the main box make SURE you removed the DFS sync partner before uninstalling DFS on the backup)

So I did what you said and with DFSR reinstalled on the 2nd server I added one replicated group (out of 7) and it's not replicating even though of course the DFS GUI says everything is fine and can replicate through topology.

This popped up again on the source server but I don't want to run that because I'm not sure the results, why would I even have to resume replication if I have no replication?



Volume: E:
GUID: EC86BB83-9428-4CFF-9E6B-71E7BDC9B7F2

Recovery Steps
1. Back up the files in all replicated folders on the volume. Failure to do so may result in data loss due to unexpected conflict resolution during the recovery of the replicated folders.
2. To resume the replication for this volume, use the WMI method ResumeReplication of the DfsrVolumeConfig class. For example, from an elevated command prompt, type the following command:
wmic /namespace:\\root\microsoftdfs path dfsrVolumeConfig where volumeGuid="EC86BB83-9428-4CFF-9E6B-71E7BDC9B7F2" call ResumeReplication

Edit: I have a backup + shadow copies going to just run it, YOLO

 

[Guldan]

Ok so that didn't work, I googled a bit and my memory refreshed.

I had top stop DFSR on both sides then go into E:\System Volume Information (After adding permissions) then delete the DFSR folder.

I restarted DFSR and created one test repl group and seems to be working so far.

 

[Biznatch]

When the replication stops due to a bad DFS service restart, you just need to go to the DFS section in event viewer, grab the WMI command and run that in elevated cmd prompt. It will force a ressync and you should be good. Although if it's not hub/spoke, any deletes that happened on the disconnected server will get resyncd once you run the WMI.

If this wasn't AD replication, I don't believe you have to worry about it being unreplicated for over 30 days. Fixing AD replication over DFS when a server has been off for a month is a pain in the ass.