All, Looking at a new exchange design where my customer has a LAN extension to another site (10Mb connection) and what I thought about doing rather than implementing a DAC setup was build the server in my 2nd site with the CAS, HT & MB roles add it to my DAG and possibly add it to my CAS array. I think this might go against MS best practices so am looking for some technical do's and dont's. My exchange design plan is to have in the production site 2 x CAS+HT and 2 x MB - have a CAS array setup using WNLB and a DAG setup with 2 live databases on 1 MB server and another 2 live databases on the other MB server with replica's of these DB's to the other MB server. My plan for the "DR" site was not to implement a DAC but rather use the LAN extension and add this "DR" server to the DAG and CAS array therefore if I have outage/DR and can easily get my users to connect to this exchange box (I do have a DC in my DR site). I am not sure on the CAS array side of things as I don't want users connecting via this server for live/production work so I guess the question is if I omit this box from my CAS array how can I get the users to connect to this device in the event of a DR senario would I just need to amend the mapi profile? Thanks

Not really sure what you are trying to achieve here using a LAN extension. If you don't seperate the servers in to AD sites, then servers in both locations will be used. The RPC CAS Array is site specific. However in the event of a fail over of the DAG, you can point the CAS array host name at the server in the other site. Therefore the simple solution is to have the two sets of servers in their own site in AD. An RPC CAS array with a short TTL time in the primary site with the Active mailboxes. In the event of complete failover, simply repoint the DNS. Simon. Simon Butler, Exchange MVP Blog | Exchange Resources | In the UK? Hire Me.

Not really sure what you are trying to achieve here using a LAN extension. If you don't seperate the servers in to AD sites, then servers in both locations will be used. The RPC CAS Array is site specific. However in the event of a fail over of the DAG, you can point the CAS array host name at the server in the other site. Therefore the simple solution is to have the two sets of servers in their own site in AD. An RPC CAS array with a short TTL time in the primary site with the Active mailboxes. In the event of complete failover, simply repoint the DNS. Simon. Simon Butler, Exchange MVP Blog | Exchange Resources | In the UK? Hire Me.