Want to protect your cyber security and still get fast solutions? Ask a secure question today.Go Premium

x
  • Status: Solved
  • Priority: Medium
  • Security: Public
  • Views: 225
  • Last Modified:

Exchange 2007 Database copy help

Hi all, we are about to get some support to help us with an Exchange Backup issue and possible corrupted Transaction Logs. Before we do this I was wondering if it was at all possible to simply dismount the Exchange Mailbox Database, then physically copy the Exchange Server\Store\Mailbox Database.edb file to a save place (spare disk or something) as a backup prior to working on Exchange? We currently have no real backup, so would like to take a copy of the store as it is.

Should everything go rather wrong, can we simple copy the saved Mailbox Database.edb back to it's original location and re-mount the store?

Many thanks,
0
georgefowler
Asked:
georgefowler
  • 3
  • 2
1 Solution
 
Andrew_CzCommented:
This article should help you with the backup & restore process (should you need to):

http://www.msexchange.org/tutorials/Configuring-Exchange-Server-2007-Preview-Chapter9-Disaster-Recovery-Exchange-Server-2007.html

Basically you use the ntbackup utility included in your server to do the backup.  Exchange is a tricky beast sometimes though and because of that, I would also do a full image of the server if possible using something like Acronis' TrueImage bootable cd.  This won't fix your current exchange issue but at least it'll put you back into the same situation you're in now so that you can try fixing your original problem again.
0
 
georgefowlerAuthor Commented:
Thank you Andrew, we've tried the NT Backup (or WIndows Server 2008 equivalent) and we are still getting the same issues where Exchange says it hasn't had a full backup since X date. Sadly it appears legitimately backing the server up using Windows Backup or Backup Exec is not possible till we fix the issue with the Transaction Logs.

I can look into the imaging method, but we are about to run into an official MS Support call, however before this it has been suggested that I take a copy to disk of the Mailbox Database, and I was asking if it is possible to dismount the database, physically copy the Mailbox Database to disk, re-mount the database again?

Appreciate the help.
0
 
Andrew_CzCommented:
Unfortunately, and I'm almost 100% sure, you can't just dismount the store and copy the files physically and if need be to try and restore them.  They won't be usable.  I've tried this myself before and I could never get it to work that way.  AFAIK using ntbackup or symantec backup exec (and other apps) are the only way.  But it seems you don't want to back it up, only to fix the transaction log issue.  If so, I'd still do a backup using Acronis before you try to fix it because you could easily make it worse.  Once you have a full backup of the server you could then mess around to your hearts content, trying various methods.

Again, you can dismount and copy but you won't be able to restore it by just copying the files back.

As for the transaction logs, I've heard before some people got it resolved by just dismounting and remounting.  This article will help you dismount it.  Then just remount it and see if you still have the same issues.

http://technet.microsoft.com/en-us/library/bb123903%28EXCHG.80%29.aspx
0
Making Bulk Changes to Active Directory

Watch this video to see how easy it is to make mass changes to Active Directory from an external text file without using complicated scripts.

 
Andrew_CzCommented:
Also, have you tried the ESEUTIL repair utility? It doesn't hurt to try.

http://technet.microsoft.com/en-us/library/aa996773%28EXCHG.65%29.aspx
0
 
lucid8Commented:
Taking the database offline and making a physical copy of it is certainly doable but rolling back to that copy depends upon what you do post backup.  so;

1. The issue you are currently having is that the backups wont complete, correct?  Whats being reported post backup in the event log i.e. why does it say its failing?

2. where you you attempting to backup to, i.e. disk, tape?

3. What is the size of the database?  What is the current log count and size?

4. Are there any other database or exchange related errors in the event log?

5. Are there any disk related errors in the event log?
0
 
lucid8Commented:
OK glad you think you have this resolved but either I completely misunderstand your post or you are missing some critical information that you need to review and I am not telling you this for points I am telling you these things based on my 29 years in the industry, 15 years with Exchange as a DR consultant and Exchange centric ISV

If you had to run an eseutil /p aka a repair on a database post dismount you have a serious issue on your hands that needs to be attended to IMMEDIATELY.    

It simply is NOT normal to have to do this and you need to realize that an eseutil /p in short scans the database looking for corrupted pages and removes them when found, i.e. data loss takes place which is not a good thing at all.  

Now if you find yourself in this position where you have a db that will not mount then /P may work for you but it should ALWAYS be the LAST option and once you do a /P against a production database you should then also do an eseutil /D the defragment the db and when that is done you should run isinteg -s <ServerName> -test allfoldertests until all errors are removed, i.e. if the first pass reports that it fixed ANY errors you need to run isinteg -s <ServerName> -test allfoldertests again and repeat until it reports that zero errors have been found.  Now you can put this DB back into production however as a best practice it is NEVER a good idea to leave a database that has had a /P against it left in production so as soon as possible you should create a new database and move the mailboxes into the new database and dispose of the old or dial-tone it once all moves are done.

But lets go back to the beginning, i.e. why oh why is a database that is being properly dismounted needing a /P?  The answer here is that A. it wasn't dismounted properly, B. perhaps it was restored from a backup to an alternate location and the logs were not rolled up to make the db consistent or C. I read this correctly and your backups are not finishing and you properly dismounted the database and had to use a /P to get that copy to function.  If that's correct then the database has been experiencing errors all along and you are on your way to a failure.  Look at my original post items # 4 and 5 and you will more than likely find out whats going on.

0

Featured Post

VIDEO: THE CONCERTO CLOUD FOR HEALTHCARE

Modern healthcare requires a modern cloud. View this brief video to understand how the Concerto Cloud for Healthcare can help your organization.

  • 3
  • 2
Tackle projects and never again get stuck behind a technical roadblock.
Join Now