Pages

Search This Blog

Saturday, April 21, 2012

Recovering a damaged Exchange 2010 DAG mailbox server

Last week I was engaged with one of our customers who was having issue with one of his Exchange 2010 mailbox servers. The issue was that the server undergoes a RAID failure which caused loss of the server installation and thus server needs to be recovered.

This customer’s mailbox server was a member of a database availability group and it was hosting a single mailbox database copy. In this post I’ll clarify the required steps to perform this action ( DAG mailbox server recovery )

Our environment will be as follow :

  • 1 Server with Exchange 2010 SP2 Hub transport & Client Access Server roles – hosting database availability group’s file share witness - , in addition of being a domain controller.
  • 2 Servers with Exchange 2010 SP2 mailbox role named Ex-01 & Ex-02 , both are member of a single database availability group named DAG.

Both mailbox servers are hosting a mailbox database named MBDB-01, as shown below:

clip_image001

We will consider that Ex-02 is the damaged server , which need to be recovered to its original state.

In order to recover the server , all related data for database availability group must be removed , as follow :

1. Any database copies were hosted on this damaged server : In order to do this we will utilize Exchange PowerShell “RemoveMailboxDatabaseCopy” CMDlet , as below :

clip_image002

2. Damage server must be removed from database availability group : Trying to do this results in a failure ( Both with EMC & EMS ]

clip_image003

clip_image004

Those failures are due to inability of cluster service to communicate with Ex-02 server ,so in order to perform a server removal from our database availability group I’ll add the –ConfigurationOnly switch to the command , as below :

clip_image005

As you can see it worked, and Ex-02 mailbox server was removed successfully database availability group

clip_image006

3. After removal from Exchange database availability group , the damaged server node must be evict from the cluster ( Using Failover Cluster Manager snap-in ) :

clip_image007

Confirm the action:

clip_image008

Now, Ex-02 node has been totally removed from database availability group cluster.

4. Reset the damaged server computer account (Using Active Directory Users & Computers snap-in)

clip_image009

5. Install and patch OS on the server , join the domain with the old name & install Exchange server mailbox role prerequisites

6. Mount Exchange 2010 SP2 DVD and from command line browse to it. At command line run “Setup.com /m:RecoverServer” , this will cause Exchange server setup to perform Active Directory discovery for each role(s) was/were installed on the server and reinstall it .

clip_image010

clip_image011

7. After rebooting , verifying installation , services and install required updates, re-add the server to database availability group :

clip_image012

8. Finally , re-add the database copy for MBDB-01 :

clip_image013

Now, you are done with the damaged Exchange 2010 database availability group mailbox server recovery clip_image014