I'm not very familiar with how DFS works, and since someone else administers the servers in question I don't know what the DFS setup looks like. I'm just wondering if something is possible.
Basically we have 3 servers setup as a replication group with 1 folder, that folder being the root of a .NET webapp setup in IIS. When we deploy changes to the application we do so to one server, Server #1, and allow DFS to propagate the changes across the group to the other two servers. There is a lot of odd behavior with the environment after deployments, where the application is taking a lot of time to come back up, but nothing in the IIS logs indicate that the app pool cycling is the part that's taking a long time. I'm thinking it may be due to odd DFS Replication behavior so I requested the admin provide me with a DFSR Health Report and Propagation Report. The Propagation report indicates quick replication, <1 sec for everything.
The Health report shows 1 error on 1 server: Due to ongoing sharing violations, DFS Replication cannot replicate files in the replicated folders listed above. This problem is affecting 114 files in 1 replicated folders. Event ID: 4302
The MS KB here: http://support.microsoft.com/kb/968429 says the 4302 event is for errors during the "Receiving" of files, which is what worries me since this error is happening on Server #1 which is what should be sending the files to the other two servers.
Is it possible that DFSR is attempting to replicate files BACK to the source server after getting one of the other two servers in sync and getting permission denied because replication to the 3rd server is still ongoing? Or is there something else I can look for in the logs to see how long replication is actually taking between the servers to try to get an accurate timeline of what happens between the end of MSDeploy and the application being available again?