Just like in Exchange 2010, Exchange 2013 has Database Availability Groups. Creating them now relies on two methods – PowerShell and the Exchange Administration Center. This article will cover the process for creating a DAG in the EAC since the 2007/2010 style GUI is no longer available.
First we need to log into the EAC:
Once logged in, you will see the main screen for Exchange Server 2013 administration:
Select Servers on the left side and this will take you to server management:
To configure a DAG, click on the “Database Availability Groups”:
At the DAG screen, click on the “+” to add a new DAG:
At the new DAG screen, give the DAG a unique name (*), Witness Server, File Share Witness directory and give the DAG an IP:
Now how do we change the DAG membership or a new DAG Network? Well…. Seems as if Microsoft did not make this obvious either. In the DAG window you see these buttons:
But, if you highlight the DAG you see some new buttons:
New DAG Network – left button
Manage Membership – right button
Now that we have a DAG we can add Exchange 2013 mailbox servers to it. To do so, select the “Manage Membership” button to add servers to the DAG:
Click on the “+” to add a new member to the DAG. Once you add the member, you will see the progress of Failover Clustering and addition to the cluster being processed:
However, with Exchange Server 2013 still being in preview code, you will get an error trying to add a member to the DAG.
The solution (which one of my colleagues discovered) is to create a create a Computer object for your DAG in AD:
After the object has been created, select the properties of the object and go to the Security tab:
Add the computer objects for the DAG members and give them full access to the DAG computer object:
Once all DAG members have been added, disable the DAG computer object:
Then retry the DAG membership addition and the error will go away. If there is a need to remove a DAG member, then you need to re-enable the same AD Computer object for the DAG and then remove the Exchange server from the DAG.
If you are going to manually add Replication Networks in Exchange 2013, you will need to turn off the automatic configuration with the ‘set-databaseavailabilitygroup -manualdagnetworkconfiguration $true’ command in PowerShell:
Select the New Network Button and add a new network:
** NOTE ** You will get an error that a server needs to be added to the DAG before this works:
Once a member has been added to the DAG you can add a copy of the database to the server.
Change to the Click on the Database tab at the top:
At the database screen click on the “+” button to add a new database copy:
Enter the information needed for the database copy:
Click Save. You may get an error when creating the database.
Lastly you will need a DNS entry for the DAG object. Open up the DNS Management console and add a new ‘A’ record.
Make sure you check the box that says “Allow any authenticated user to update DNS record with the same owner name”. Also, change the TTL from 0:1:0:0 to 0:0:5:0 which changes the DNS TTL to 5 min from 1 hour. http://technet.microsoft.com/en-us/library/dd638104.aspx
Finish this up by clicking ‘Add Host’.
Now that your DAG is created you can enjoy High Availability built into Exchange Server 2013.
As noted in this article by Microsoft, the DAG name must be unique. Now how unique does it have to be? In the domain, in DNS, or on your entire network? Well, as it turns out, it must be unique on the entire network. My test lab is connected to the same network backbone as my company’s. This means that I can ping and access company servers from the lab. This proves useful for accessing IS files as well as other software for the lab. However, as noted by the unique name requirement, this can cause issues if you name you DAG with a name of a DAG already on the network.
If you do create a DAG with the same name as one already on the network, you can still add members and database copies. However, you will have client connection issues as well as DAG management issues.
This is how you can tell you have an issue. Open up Failover Cluster Manager and you will see that your DAG has a red mark on it. When you review you cluster resources, you will see the following:
If you review the cluster events, there are two events that occur:
Just keep this in mind when creating a new DAG either in a lab or production environment.