Hello, I've done a fair amount of research on the topic (including a lot of great posts here), but I'm in a bit of pickle with a deployment. The customer is demanding a level of automation when there is a failure in either site. Currently all three members are in the primary site, but one member will be moved shortly to the DR site. I will have the FSW and two members of the DAG at the primary site and the third member in DR with the aFSW. DR is a site with users and CAS/HT/MBX roles are installed on all servers. We have Kemp HLB's at the primary site and intend on using a CAS array in the DR site. The conversation went south when I began to explain what happens when DR is cut off from the primary site or if the primary site were to lose power. They want to exhaust all possible scenarios that do not involve making manual changes. Is it possible to switch the DAG from node majority to disk only and reference a DSF-R share between the two sites and what would be the major drawback with this design? Since users are located in DR site, would it be recommended to house them on database that are mounted locally? If the CAS array in the DR site goes down, I believe that users there would reference the ExternalURL setting for outlook anywhere. Is my understanding correct here and how long will it take the clients to timeout and use the ExternalURL setting? Is anyone automating the primary datacenter recovery using DAC? i.e. If power is lost in the primary site, so no pre-emptive changes would be made. When power is returned, will the servers check for the DACP bit automatically and then replay the logs from the DR site or does something need to take place before logs are replayed. I looked at the technet article on recovering the datacenter, but I'm still a little confused on that part. I understand that once the DB copies are healthy, a maintenance window is recommended to mount them in the primary site. Is there any major risk to moving the active copy back to their original location outside of a maintenance window and, if not, automating that process. Thanks! -skint

Hi Skint, I would suggest that you could get some information from below DOC firstly and carefully: http://technet.microsoft.com/en-us/library/dd638121.aspx Per your plan, I do not think it could achieve your target. 1. You want to deploy one server in the DR and hosts the cas/hub/mailbox together, I would not do so, instead install the DAG memeber server individually. 2. If you just deploy three memebers in the DAG and let only one member server in the DR, and if the primary DATACENTER failed, the whole email system will not achieve the site resilience, because of that more than half member server failed would cause the DAG failed. 3. It is better to deploy four memver server, like below referred: http://technet.microsoft.com/en-us/library/dd979781.aspx 4. If you want to use the HLB, per my known, some third party product support the cross site load balance, you could do some research about it. 5. And if after failover between the DAG member server, the active node would not automatically were changed on the legacy server, you could refer to the DOC on the Technet. Regards! Gavin Please remember to click Mark as Answer on the post that helps you, and to click Unmark as Answer if a marked post does not actually answer your question. This can be beneficial to other community members reading the thread.