cancel
Showing results for 
Search instead for 
Did you mean: 

Exchange 2010 Journaling

Fremont
Level 5
Partner Accredited

Hi,

I am in the process of designing EV9.0 for journaling with Exchange 2010 and have a number of questions.  There is a white paper that explains the archiving side as well as Journaling IRM and report decryption, but I have not seen much info on the journaling design side.  The questions revolve about the use of DAG with journaling:

- Is it recommended to put the exchange journaling mailbox in a DAG? 

- If yes, do we create journaling tasks to point at all DAG member servers just like for archiving tasks? 

- If we don't, does journaling stop or continue after DAG failover for the DB that houses the exchange journaling mailbox? 

- If it continues, how does it go about it?

 

I know some of these are really Exchange questions, but I thought I'd post since this is so new that we could all benefit from each others' experiences.

Thanks,

1 ACCEPTED SOLUTION

Accepted Solutions

KarlW
Level 6
Employee

Hi

There is no reason why a journal mailbox cannot be on a database being copied within a DAG.  If you do so there is no need to create journaling tasks pointing to all DAG member servers, the one task is sufficient (the journal mailbox when created as a target is linked to a task and doesn't move on failover).

Journaling should continue on failover of the DB.  You may see a small interruption of the journaling task if it is in the middle of processing a journal mailbox or message as it will receive an error from Exchange.  This will result in a 1 minute pause of the task.  After which hopefully the DB has successfully been mounted on another member of the DAG and the journal task will continue.

The journal task really doesn't have any specific logic to continue (other than recreating MAPI sessions and re-opening the mailbox) as the Exchange Client Access Server (or array) and DAG handles the DB failover and makes sure MAPI traffic talks to the correct DAG member.

Best practice would be to make sure all RPC Client Access Servers for the databases are really a CAS array.

Regards

Karl

View solution in original post

2 REPLIES 2

KarlW
Level 6
Employee

Hi

There is no reason why a journal mailbox cannot be on a database being copied within a DAG.  If you do so there is no need to create journaling tasks pointing to all DAG member servers, the one task is sufficient (the journal mailbox when created as a target is linked to a task and doesn't move on failover).

Journaling should continue on failover of the DB.  You may see a small interruption of the journaling task if it is in the middle of processing a journal mailbox or message as it will receive an error from Exchange.  This will result in a 1 minute pause of the task.  After which hopefully the DB has successfully been mounted on another member of the DAG and the journal task will continue.

The journal task really doesn't have any specific logic to continue (other than recreating MAPI sessions and re-opening the mailbox) as the Exchange Client Access Server (or array) and DAG handles the DB failover and makes sure MAPI traffic talks to the correct DAG member.

Best practice would be to make sure all RPC Client Access Servers for the databases are really a CAS array.

Regards

Karl

AndrewB
Moderator
Moderator
Partner    VIP    Accredited

hi fremont, i'm curious to know what you ended up doing for this. please share with the forums if you're still around.