Hello, We have one doubt about the best implementation of DFS Replication to solve the exponential growth of data that our company has experimented.Now I have two branch office (BARCELONA and MADRID) and we have one storage server to support the current work files. The two servers are twin and they have 1.36 TB mounted with five array level and they are replicating with mesh topology with a 24 h window. The servers are reaching the limit of the storage capacity and we want to implement the best scalability and performance solution .Our doubt is the next information about the tested scalability about DFS Replication, realized by Microsoft engineers:A volume can contain up to8 million replicated files, and a server can contain up to 1terabyte of replicated files.What do they want to say with "a server can contain up to 1 TB of replication files"? That the server that contain the root of DFS only can to manage 1 TB of replication files so is it possible to the server use one LUN of 1 TB and other server other LUN of 1 TB, both located into the same SAN (ISCSI connection) or that the physical machine that storage the files only can have 1 TB of available space.I would like to know that is the best:To put two servers. One server in BARCELONA and the other server in MADRID. This solution want to say that next year we are having two servers more, and gowing ... but we will have one server more to backup every main storage server.To put two SAN (One in BARCELONA and the other in MADRID) with the storage size required in this moment, put two servers and manage the new LUN, and put the data of old servers into other LUN and map this LUN to this server. The old server and new server of every branch office manage the replication files of our mapped LUN and the their internal storage can be used to backup its mapped LUN.The solution two are involving than SAN was storaging more than 1 TB of information that it is replicating with DFS Replication technology.Thank you very much for helping. Ivan Fraixedes

Ivan Fraixedes said: Our doubt is the next information about the tested scalability about DFS Replication, realized by Microsoft engineers: A volume can contain up to8 million replicated files, and a server can contain up to 1terabyte of replicated files.What do they want to say with "a server can contain up to 1 TB of replication files"? A: Actually, this is not a hard limit on the number of files that can be replicated. For detailed information, you may find the explanation from the DFS Replication program manager via the following link (Microsoft Storage Team Blog) More on DFS Replication Limits http://blogs.technet.com/filecab/archive/2006/02/09/more-on-dfs-replication-limits.aspx Understanding DFS Replication "limits" https://blogs.technet.com/filecab/archive/2005/12/12/415942.aspx That the server that contain the root of DFS only can to manage 1 TB of replication files so is it possible to the server use one LUN of 1 TB and other server other LUN of 1 TB, both located into the same SAN (ISCSI connection) or that the physical machine that storage the files only can have 1 TB of available space. A: Theoretically, it is possible for us to make Server1 and Server2 (which are located in two different sites) use LUN1 and LUN2 which belongs to the same SAN. However, it is not recommend that you perform like this. Because iSCSI SAN depends on the TCP/IP network, you may need to verify that all the network devices between the 2 sites (BARCELONA and MADRID) support iSCSI protocol. Meanwhile you may also confirm that the all the iSCSI related port is not blocked on the firewall between the 2 site. Moreover, we need to ensure the network stability to make the servers connect to the LUN continuously. In this way, for the consideration of business consistency, if possible, I'd suggest that you implement your second proposal. To put two SAN (One in BARCELONA and the other in MADRID) with the storage size required in this moment, put two servers and manage the new LUN, and put the data of old servers into other LUN and map this LUN to this server. The old server and new server of every branch office manage the replication files of our mapped LUN and the their internal storage can be used to backup its mapped LUN.Hope the explanation will be helpful. David Shen - MSFT