Windows 2012r2 Server DFS Questions

Am I correct in my understanding (I’ve done a bit of reading on DFS) that each server will contain all of the shares (via replication) and that if one goes down the shares will automatically be available from the other server through DFS? How does that look in a typical drive mapping (we still use login scripts).

We were hoping to setup two storage servers in our two offices using Windows Server 2012 R2 with DFS
Who is Participating?

[Product update] Infrastructure Analysis Tool is now available with Business Accounts.Learn More

I wear a lot of hats...

"The solutions and answers provided on Experts Exchange have been extremely helpful to me over the last few years. I wear a lot of hats - Developer, Database Administrator, Help Desk, etc., so I know a lot of things but not a lot about one thing. Experts Exchange gives me answers from people who do know a lot about one thing, in a easy to use platform." -Todd S.

Cliff GaliherCommented:
DFS-R isn't meant to provide high availability. If you try, you'll end up with conflicts where person A edited a copy on server A while person B was editing a file on server B. This is by far the biggest mistake and most misunderstood part of DFS-R. By orders of magnitude. it sounds like you want a file server cluster. That's still the way to provide HA for file shares.
JesusFreak42Author Commented:
Would we be able to do this with a server in each location?
Cliff GaliherCommented:
That depends on your infrastructure.
Acronis True Image 2019 just released!

Create a reliable backup. Make sure you always have dependable copies of your data so you can restore your entire system or individual files.

JesusFreak42Author Commented:
Sorry, I don't know a lot about FIle Server Clustering specifically. But basically here's what we have:
Each Location has:
- 1 Domain Controller/DNS
- 1 SQL Server
- 1 Application Server

One location has a couple NAS. We thought we could consolidate the NAS' into one server at one location and then install and identical server in another location and use DFS. You seem to be saying that this is a bad idea, and that it does not give real redundancy or help with file conflicts. So I guess we need to look at a different strategy.
Cliff GaliherCommented:
It *does* provide redundancy. But so do backups! Redundancy is not the same as high availability. And HA takes proper planning. Used properly, DFS-R can help in centralizing data for later backup or for making access to large files in a remote location much easier. But it does not replace clustered file servers and attempting to use it *that* way will cause version conflicts. So whether DFS-R is a good fit depends on the task and the usage patterns.

Based on your description, I still don't thin DFS-R is what you want. DFS-N can certainly help you consolidate your NAS (you may not even need DFS-N)), but the "identical server in another location" is a new addition, not a replacement or consolidation, and doesn't sound like a good fit for DFS-R. Since I doubt your NAS boxes were highly available, you may not need a cluster either. A simple replacement file server would meet your goals of consolidation.
JesusFreak42Author Commented:
Ok. Just for clarification, what WOULD happen if the main file share went down and we were using DFS? would the other one still be available?
Cliff GaliherCommented:
That depends entirely on how you set it up. DFS-R and DFS-N have no inherent codependence. You can deploy either one without the other. You can replicate data without the namespace being shared and thus no, if server A went down, any shares it hosted would be inaccessible until you brought it up.
JesusFreak42Author Commented:
So to sum up
DFS-R = Replication with no redundancy
DFS-N = Replication of Namespace only
Both = Replication of both, resulting in redundancy, but versioning conflicts.

Two more questions and then I'll start a new topic because you've been so helpful:
1) If there is LIGHT traffic, but the need for redundancy just in case, would DFS N+R work well? I guess I am trying to figure out when to use it at all.
2) Any good articles on setting up file clusters?
Cliff GaliherCommented:
No, that isn't how I'd describe any of that.

DFS-N allows you to build a common namespace across multiple servers (take many servers, make them appear as one.)  There is no replication of data except insofar as they are in AD and AD replicates across domain controllers.

DFS-R replicates data. By definition, that does offer redundancy of *data.* But there is no guarantee of redundancy for connectivity.

Both: If DFS-N defined multiple targets and DFS-R replicated those same targets then yes, you'd get the appearance of redundant connectivity and data. There is no requirement for DFS-N and DFS-R to reference the same folders, however. You could use DFS-N and DFS-R disparately as well making some very cool, yet complex topologies.  Thus "both" is an overly simplistic representation that may not necessarily even provide the appearance of redundancy. I say the appearance of, because there are other factors (link speed, link redundancy itself, replication interval, amount of data being replicated, etc) that can impact this. The redundancy may exist, but is neither simultaneous nor guaranteed, and yes versioning conflicts can result if the workflow does not account for this.

1) I would argue that no, the amount of traffic isn't the issue. It is *how* the system is used.
2) Plenty. Start with TechNet. If you need more specific help, then ask.

As far as when to use example of where it works well:

An engineering firm has field employees that do large site surveys and load up files into the server in a field office.  From there DFS-R replicates the site surveys to the home office. The home office can now back up all field office data.  Once a site survey is complete, the surveyor informs the legal department. The legal department reviews, edits, and signs off on the site survey. Those changes replicate back down to the field office. Once legal signs off on the project, the local contractor at the field office can open the project file and begin work based on the feedback from legal.

In the above instance, version conflicts will not arise, not because of any technical blocks, but because of how work is performed. Maybe a product like Microsoft Project is used to orchestrate this workflow. But the point is the site survey folks only work on the file until they hand it off to legal. Then *only* legal works on it. Then only the contractors work on it. Because of their workflow (purely business oriented, not technical) there is no reasonable way that a file can get edited at both the site location and the home office at the same time. DFS-R only replicates between the site and the home office. If HA were required, a file-cluster would have been set up. So to expand our example, the home office may, in fact, run clustered file servers. And DFS-R replicates from the cluster to the satellite offices. That is perfectly acceptable. You *still* don't have two people editing two different versions on opposite ends of the DFS-R link.

Make more sense now?

Experts Exchange Solution brought to you by

Your issues matter to us.

Facing a tech roadblock? Get the help and guidance you need from experienced professionals who care. Ask your question anytime, anywhere, with no hassle.

Start your 7-day free trial
JesusFreak42Author Commented:
You have been EXTREMELY helpful to a networking guy trying to understand the way this MS OS thing works. :)
It's more than this solution.Get answers and train to solve all your tech problems - anytime, anywhere.Try it for free Edge Out The Competitionfor your dream job with proven skills and certifications.Get started today Stand Outas the employee with proven skills.Start learning today for free Move Your Career Forwardwith certification training in the latest technologies.Start your trial today
Microsoft Server OS

From novice to tech pro — start learning today.