asked on

Recovering exchange DB file after crash in HDD RAID0 failure

Hi there,

I´m struggling with an issue on my exchange server 2007 regarding a database that is in a dirty state. the problem is that this DB file is on a server with a RAID 0 and one of the disks is failing. I´ve managed to force the disk to "malfunction" in order to be able to get the file from it. If I manage to get a copy of the file (db.edb) (the log files are in another drive) will it be possible after rebuilding the RAID with another disk and restoring the volume name and file to access the exchange console and do an ESEUTIL to try to repair the excahnge DB? how can I recover my mailboxes? this is URGENT as we only work with emails and as I write this everything is stopped

ASKER CERTIFIED SOLUTION

M A

membership

This solution is only available to members.

To access this solution, you must be a member of Experts Exchange.

Start Free Trial

Mal Osborne

RAID 0, also known as striping, is anti fault tolerant. Assuming drives that have 90% lifetime failure rate, a RAID 0 of 4 drives would have a 66% chance of loosing data, while the same drives in RAID 1 would be have .01%. RAID 0 is usually only configured for data that you do not mind loosing.

One failing drive means that data is already inaccessible by definition. IF it is only a few sectors not reading, there are a few options you have.

1. Replace the drive, then restore from backups.
2. Install another drive, at least twice the capacity of the other two, and see if you can get data to copy over to it. Once you have moved the data, then see if ESEUTIL can help make sense of it and let you mount the store.
3. If the data has value, and there is no backup, then send the box to a data recovery centre, who have the expertise to disassemble and recover data from broken drives. These services cost thousands, and are not always successful.

It would be a big mistake to run ESEUTIL against a RAID0 set with a broken drive, you would almost certainly make matters worse.

Windindi

ASKER

I use the exchange troubleshooting assistant and it did a ESEUTIL /p and the result was this:

Unknown Database (Mailbox Database) finished the database repair action with positive status code Mailbox Database Eseutil.exe /p finished action on database (Mailbox Database) with status code as 595

what can I do it seems that it did not repair the error

Windindi

ASKER

did it again and I get this:

Analysis results: Mailbox Database
Database status is 'Clean Shutdown'

Windindi

ASKER

But still not able to mount the Database

Windindi

ASKER

After trying to defrag again I got this error:

Operation terminated with error -1605 (JET_errKeyDuplicate, Illegal duplicate ke y) after 2.278 seconds.

Windindi

ASKER

If I copy all ost files from each PC (I also exported the ost to pst files) can it be possible to do a scratch DB to connect the exchange and then feed those ost data back ?

Windindi

ASKER

Would this be a solution to my problem?

I´ve copied all OST to a hdd and purchased Stellar software (was unable to recover the edb file with the info that it is severely damaged) so now I´m building pst files from the ost info. Now since I don´t have any edb file but I´m still able t see the mailboxes on the EMC. Could I create a new storage group different from this first storage group and create a new DB and start moving the mailboxes and then fill it with the pst info?

Windindi

ASKER

Thanks MAS for the help... I was working with a edb that was really damaged so went for a backup and all worked fine and did not even had to do an ESEUTIL /P just ESEUTIL /mh to the logs and edb and then with the exchange assistant mount the DB and all worked fine automatically,

best regards your links helped a lot

M A

Glad to know you fixed and pleasure helping you again. :)