Hi All, Please could you tell me why is offline defarg no londer required for Exchange 2010 database? I read this article http://technet.microsoft.com/en-us/library/bb125040.aspx#NewESE, but i did not understand about offline defarg. Also, how do I run esetuil for DAG envirionment? Regards.Regards. Jos Osorio.

Why do you want to run an offline defrag? If you have high amounts of white space, create a new database and then move all the data to the new database and drop the original. If you need to repair a database in a DAG then you will have to take the database out of the DAG, then run the repair and put it back in the DAG and reseed it. However that may not be necessary - as the corruption is unlikely to have replicated to the DAG, so if you have a bad database then you can fail-over and reseed. Simon.Simon Butler, Exchange MVP Blog | Exchange Resources | In the UK? Hire Me.

Offline Defrags were never required or recommended as standard maintenance for any version of Exchange.

Hi Simon, What do you mean When you said " take the database out of the DAG"? Do you mean to remove the Mailbox member from the DAG?, because just Mailbox servers are menbers of DAG. I think this : 1. Suspend database Copy. 2. Dismount Active Database Copy (When I do this. Does Passive Copy Database become Online?) 3. Run Eseutil. 4. Mount Active Databae Copy 5. Reseed. Is that right? Regards. José OsorioRegards. Jos Osorio.

So, why is Offline Defarg often required?Regards. Jos Osorio.

I would avoid an offline defrag, there is no reason to ever do it for the most part.

When do you say it's required? Agree with Andy, it has never been required or recomended. lasse at humandata dot se, http://anewmessagehasarrived.blogspot.com

What I mean is that if Offline Defarg is not requiered. Why there is a parameter in eseutil command to defrag exchange databse? or In which case is requiered? Regards.Regards. Jos Osorio.

eseutil was not developed for the Exchange admin in mind, it was developed for the ESE dabase people at MSFT to use. Because there is an option does not mean you got to use it or it's good to use. lasse at humandata dot se, http://anewmessagehasarrived.blogspot.com

Hi Lasse, So the best option instead of using offline defarg is to create a new database and then move all the data to the new database and drop the original, right? Is there an exceptions when you must use offline defarg with eseutil? Regards.Regards. Jos Osorio.

If I was running standard Exchange 2000 and was running out of space on my 20GB Info Store Drive, I may run an offline defrag if I couldnt bring another server up.

Yes, this will almost be unnoticed by users compared taking a database offline and use eseutil on it for many hours. Can't come up with a reson to use eseutil. lasse at humandata dot se, http://anewmessagehasarrived.blogspot.com

Space was far more precious back in the days than it is now. lasse at humandata dot se, http://anewmessagehasarrived.blogspot.com

Ok, I understand. Please when You have to repair a Database which has an active database copy and a passive database copy. What is the procedure to repair it? I think this : 1. Suspend database Copy. 2. Dismount Active Database Copy (When I do this. Does Passive Copy Database become Online?) 3. Run Eseutil. 4. Mount Active Databae Copy 5. Reseed. Is that right?Regards. Jos Osorio.

You don't repair databases. Do a reseed instead. But if you want to, you must repair all copies of the databases, not just the active one. lasse at humandata dot se, http://anewmessagehasarrived.blogspot.com

Hi Lasse, But what happen when a database which has two passive database copy become corrupted, will not be better than repair the active and then reseed? Regards.Regards. Jos Osorio.

You shouldnt be repairing databases. That is the point of the DAG. If anything restore from backup. If all options fail , you have no good copies and no good backups, then the same principles apply whether the store is part of a DAG or not. If you dismount the store, it doesnt failover to another server, it goes offline. But yes, remove all the passive copies, dismount the store to take it offline, run eseutil, bring the store back online ,- MOVE mailboxes to another store ASAP. Do not run a repaired store in production.

Move the active database copy yo one of the passive ones, and ressed a new fresh copy to the corrupt copy. lasse at humandata dot se, http://anewmessagehasarrived.blogspot.com

Hi AndyD, I think that if the active database copy becomes corrupted this database makes all passive database copies corrupted, right? So, the only option I have is recover it from Backup or run the New-MailboxRepairRequest.ps1, right?Regards. Jos Osorio.

Not necceasry. If we are talking about logical corruption then yes it will replicate but this will not make the database go offline or something really bad and is solved with mailboxrapairrequest. More serious corruption is caused by faulty hardware and is often causing bits on disk being changed. These are not replicated since replication has already taken place. Databaes can still be running or may cause a failover when this happens. Reseed from a good copy to replace the bad copy is the solution here. Some bit errors can be corrected automatically when you have more than 3 copies of the database. lasse at humandata dot se, http://anewmessagehasarrived.blogspot.com

Hi AndyD, I think that if the active database copy becomes corrupted this database makes all passive database copies corrupted, right? So, the only option I have is recover it from Backup or run the New-MailboxRepairRequest.ps1, right? Regards. José Osorio. It depends. The storage on active copy could go bad, that store gets corrupted, but only physically. It crashes, fails-over to another server which becomes active and you reseed the corrupted copy once the storage issue is resolved. New-MailRepairRequest is for logical corruption - not physical - replacing isinteg. In that case, you you can fix at the mailbox level rather than the store level. Its important to remember that 2010 is built for HA. Corruption of the store is a rare occurence now.