I have a Windows Server 2012 R2 DFS node that got its database corrupted. Before the DFS-R corruption, it had a totally up to date data set. It was around a week before we discovered the corruption, which would have meant that around 10,000 files were changed during that week.
Fortunately I have another 2012 R2 DFS node that has a perfectly intact database, so I thought "Great! I'll just export its database, and import it to the other server!", as per the "Use cloning to replace a corrupted DFS Replication database" section of https://technet.microsoft.com/en-us/library/dn482443.aspx?f=255&MSPPError=-2147217396
So I removed the faulty server from the DFS-R replication group.
On the working server, I ran:
Export-DfsrClone -Volume G: -Path C:\Temp\DFSR-Export\
And on the faulty server, I ran:
Stop-Service DFSR
Remove-Item –path "G:\system volume information\dfsr" –recurse -force
Start-Service DFSR
Import-DfsrClone -Volume G: -Path C:\Temp\DFSR-Export\
I then re-added that server to the replication group, and then after giving it an hour or so to get itself in order, I ran:
Get-DfsReplicationGroup APPS | Get-DfsrBacklog -SendingMember goodserver -ReceivingMember badserver
VERBOSE: The replicated folder has a backlog of files. Replicated folder: "APPS". Count: 4242345
Wondering if it was just rehashing the files, I let it run, but it copied a full 40GB of files into the Conflicted and Deleted folder, and grabbed new copies of the identical files.
Is there any way I can troubleshoot this? I don't want to suck down almost 800GB of files over this remote link if I don't have to. As I said, the DFS volume was totally up to date and working fine until a week ago.
