I am in the process of recovering a MS Exchange 2007 server from a failed hard drive (one of 5 in a RAID 5 Array). In the process of failing, the drive actually damaged the array. I will have to delete the array, recreate it and restore from backup.
My problem is that in repairing the volume, chkdsk was able to "repair" an unloadable mailbox database by replacing damaged clusters. This has given me a database that will load and operate, but one that I can't backup.
I am thinking about using the essutil to repair the database, but it is 100Gb in size. From my estimates, this would require 40+ hours to repair.
I have been told that I should not let the database get that large, and that I should split it into multiple stores.
If I create a few new stores, and transfer the mailboxes across those stores until my single store is empty....
1. Can I delete the damaged store?
2. Will the process of moving mailboxes remove the errors and perform the defrag one mailbox at a time?