We are setting up 3 replication groups, 1 which we are able to do over the LAN and the other 2 which need to be performed over the WAN (10mb IPSec VPN at each site).
The local data set which is 2TB replicated fine in about 2 days.
The 2 WAN data sets have exhibited strange behaviour. They start off quick, then basically hit a block and continue to replicate but very slowly (500mb-1gb/day).
The first one got up to 220gb, then slowed right down. The second one got up to 45gb, then slowed right down.
Hub: Server 2008 R2
Client: Server 2012 R2
The staging quota should be OK as I am not getting any errors in the DFSR log about hitting the high watermark.
If I look at the logs in C:\Windows\debug I constantly see this error:
+[Error:9027(0x2343) Meet::InstallStep meet.cpp:1879 12296 C A failure was reported by the remote partner]
+[Error:9027(0x2343) Meet::Download meet.cpp:2296 12296 C A failure was reported by the remote partner]
+[Error:9027(0x2343) InConnection::TransportRdcGet inconnection.cpp:4423 12296 C A failure was reported by the remote partner]
+[Error:9027(0x2343) DownstreamTransport::RdcGet downstreamtransport.cpp:5265 12296 C A failure was reported by the remote partner]
+[Error:9027(0x2343) RpcFinalizeContext downstreamtransport.cpp:1147 12296 C A failure was reported by the remote partner]
+[Error:9027(0x2343) DownstreamTransport::RdcGet downstreamtransport.cpp:5192 12296 C A failure was reported by the remote partner]
+[Error:9078(0x2376) DownstreamTransport::RdcGet downstreamtransport.cpp:5192 12296 C All server file transfer contexts are currently busy]
I have had a look through http://blogs.technet.com/b/askds/archive/2007/10/05/top-10-common-causes-of-slow-replication-with-dfsr.aspx however none of the points seem to apply.
I've noted this solution also, that also suggests possible VPN issues http://social.technet.microsoft.com/Forums/windowsserver/en-US/eeaa60c7-0480-4ae9-b367-4cb9676fabd0/dfsr-event-5014-dfs-replication-service-is-stopping-communication?forum=winserverDS - I know it references Meraki's VPN setup (which funnily enough we are moving to shortly, but we're just using Cisco IPSec VPN right now via Cisco Routers), but perhaps that may also have some sort of idea as to what the issue may be?
I'm inclined to think though it's definitely something relating to RPC (happy to be proven wrong though!) as if I copy the same data over the WAN using robocopy or even Explorer, it can do 550gb no problems in 3 days.