by March 10, 2007 0 comments



In present day set up, users in branch offices in an enterprise need access
to shared folders and files stored in another branch or a data center located at
some other geographic location. This process has been made simpler in Windows
2003 Server R2 with the introduction of various new technologies; one of them is
DFS Replication, which we will talk about in this article.

DFS Replication is the successor of DFS Windows 2000 Server’s FRS (File
Replication Service). In Windows 2003 R2, DFS Replication is an engine that does
replication scheduling and bandwidth throttling. It uses RDC (Remote
Differential Compression) algorithm. RDC is a client-server protocol that
updates files over a limited-bandwidth network and can detect insertions,
re-arrangements of data in files and in turn enables DFS Replication to
replicate only the changed file blocks when the files are updated.

Direct Hit!

Applies To: Network managers
USP: Using DFS replication to optimize file sharing across
branch offices
Primary Link:
www.microsoft.com/DFS

Google Keywords: DFS Replication, Windows Server

Associated technologies
Another feature called Cross-file RDC, contained in the same R2 pack, reduces
the amount of bandwidth required to replicate new files. There is also an
enhanced feature with it called as the DFS Namespaces. It was earlier introduced
in Windows 2003 Server SP1 as Distributed File System technology. It allows
administrators to group shared folders located on different servers and present
them to users as a virtual tree of folders known as a namespace. In its new
enhancements, it has processes like Client Failover in which clients can access
another server, after one of the servers has failed in the namespace. Another
process is Target Priority in which a client accesses a namespace; the client
receives a referral that contains a list of targets associated with the
namespace root or folder. To fine-tune how particular targets are ordered,
administrators can specify whether a server appears first or last in a referral.

DFS Namespaces and DFS Replication together are used to publish documents,
software, and data to users in an enterprise. Even though DFS Replication alone
can be used for distributing data, using DFS Namespaces allows administrators to
configure the namespace so that a folder in the namespace can be hosted by
multiple servers. When users browse namespace, they see a single folder and are
not aware that the folder is hosted by multiple servers. When a change occurs on
one folder, they are replicated to all members of the replication group. The RDC
component detects changes on the volume by continuously monitoring the update
sequence number (USN) journal, and DFS Replication replicates changes only after
the file is closed. Also in DFS Replication when a file is changed, only the
changed blocks are replicated, not the entire file and RDC protocol determines
the altered file blocks and transfers only the changed fraction of the file over
the network. Each DFS server can support a maximum of 256 replication groups and
each of these groups can contain up to 256 replicated folders. You can configure
one or more sets of folders and servers as a replication group with a common
configuration for replicated folders, replication schedule, and bandwidth
throttling.

In R2 you can create DFS
health report anytime to find out status of replication servers as well as
details of errors and warnings given by the servers

Using DFS snap-in present in R2, administrators can easily add new servers
and pre-stage the replicated folders on the target servers by copying the data
to the servers or restoring a backup etc. If the files on the target server are
out-of-date, DFS Replication will use RDC to replicate only the changes that
occurred since the data was pre-staged. If any pre-staged files or folders exist
on the target servers, which are not present on the source server, they are
moved to the ‘PreExisting’ folder under the target server’s replicated folder
path. In DFS Replication all global configurations are stored in Active
Directory (AD) and are also cached in a local .xml file on each member. If a
member crashes or the file becomes corrupted, DFS Replication can easily rebuild
the file using the settings stored in AD.

Implement DFS Replication
First thing you need to do is to install DFS Management Component and DFS
Replication Service. For this install Open DFS Management from the
Administrative tools to open DFS Management console. To create a new namespace,
right-click on the Namespaces option and select New Namespace. A wizard will
start; first it will ask you to provide the Name of Server, where namespace will
be hosted, and then you need to give a name to Namespace you are creating, let’s
say ‘dfstest’. Now it will ask whether you want your Namespace to be domain
based or standalone; here choose Domain and go to create Namespace, confirm the
settings and close the wizard. Now expand the namespaces and select just created
namespace i.e. dfstest and go to Namespace Servers tab. Here from the Actions
panel, click Add Namespace Server. In Namespace server field, provide the name
of another server to host the namespace and click ‘Yes’ when asked to start the
DFS service.

Now to create a folder in namespace with target in DFS Management console,
click New Folder and provide it a name. Next go to this just created folder
again and click New Folder. Provide it a name lets say ‘test’ and select Add to
add a folder target. Click Browse to browse for Shared Folders and in next step
provide the name of the server that will host this ‘test’ shared folder. Now, to
enable DFS Replication on the ‘test’ folder, from the DFS management console,
right-click on the ‘test’ folder, and then click Replicate Folder. This will
start the Replicate Folder Wizard, here it will first ask you to provide a name
for Replication Group, Replicated Folder Name and Replication Eligibility.
Select defaults. Next, in primary member window, choose any member, if the
folder targets are empty. And if both folder targets contain content, choose the
member that has the most up-to-date content. In the Topology Selection window
Select Full Mesh and in Replication Group Schedule and Bandwidth option choose
Replicate continuously using the specified bandwidth. Lastly click on Create the
Replication group and finish the wizard. A new replication group should now
appear in Replication option in the console tree. Now if you want you can try
browsing the namespace to see if it’s working properly.

No Comments so far

Jump into a conversation

No Comments Yet!

You can be the one to start a conversation.

<