1

One week before in our HP PROLIANT ML370 G5 server with RAID5 one hard disk turned amber and we removed and it is working under degraded state until now.

Yesterday morning our users in the domain complained that mails are not going from their OUTLOOK connected to our sbs 2003 exchange .All the users are configured with psts. not osts .After checking the ESM it was found that all the mails were in queue and internet connectivity and smtp smart host all reachable.So the exchange services were restarted, at that we found that the information store is starting at last moment it stops with error 1068.

We restarted the server, it took a long time to reach the desktop and it is found that the information store is not started.What is observed is when the information store service is started the server totally hangs ; we had to make hard reset the server a couple of times when we are troubleshooting.

After searching in internet we found that the information store edb files corruption leads to this symptom from the link http://support.microsoft.com/kb/313184

and we are now repairing the database as per this document ; it crossed priv1 .edb successfully but for the past 4 hours pub1.edb and stm scanning database catalog is going on.

Please advise me how to proceed further since I have to make it ready tommorrow morning this production server

N.B: the database priv1.edb and its related files - 9 GB & pub1.edb and stm files -2 GB

Thanks & Regards

S.Swaminathan


1
  • First off, did you have a backup of the database that you could have restored? And second, did you take a backup before you started recovery steps?
    – DanBig
    Feb 6, 2012 at 16:42

2 Answers 2

2

It sounds to me like you have hardware problems (likely more errors in the already-degraded RAID).

The Information Store service isn't going to make the server "totally hang", no matter how "corrupt" the database files are. If the machine is actually hanging when trying to mount the Information Store you're not going to get a successful repair running on that hardware.

Even if you did manage to repair the database it's not a good idea to continue running a production Exchange installation from a database that has been repaired. If you manage to repair it you'd be best off moving the mailboxes off the server to another Exchange instance and throwing that repaired database away.

At this point, I'd be performing a disaster recovery installation of Exhcange, restoring your most recent backup of Exchange onto a new server (virtual machine, etc), and getting away from the failing hardware that likely caused your problems to begin with. Continuing to provide a production service on hardware that has already caused downtime and data corruption is a bad idea.

If you don't have a backup then, at least, I hope you backed-up the database files before you attempted repairing them. You could try repairing those files on another machine (Exchange doesn't have to be installed for ESEUTIL to work-- a copy of the "Bin" directory from Exchange will do to supply the necessary files to run it).

If you have no backups at all then you may have lost your Public Information Store (pub1.edb and pub1.stm). Since yor say your Private Information Store did repair successfully (priv1.edb and priv1.stm) you may yet be able to get user mailbox data back.

1
1

Check out the Disaster Recovery section of this link.

http://www.msexchange.org/tutorials/exchange-isinteg-eseutil.html

You will need to run ISINTEG at the end to ensure you had a good repair with ESEUTIL.

1
  • Hi DanBig Thanks a lot for your immediate reply and the problem is solved by myself and the link is at social.technet.microsoft.com/Forums/en-US/exchangesvrgeneral/… Feb 8, 2012 at 10:28

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .