Granular Recovery Technology (GRT) based backup of Exchange Database to a Backup to Disk (B2D) Folder fails with error "Final error: 0xe00084af - The directory or file was not found, or could not be accessed"
Final error: 0xe00084af - The directory or file was not found, or could not be accessed.
When the backup fails with the error above, Beremote debug from the media server shows the following during hard recovery:
 09/02/08 00:03:36 JetExternalRestore(2) - ERROR::RC = -1018 - -1018 There is a checksum error on a database page.
 09/02/08 00:03:36 Using EDB Provider for Browse/Backup.
 09/02/08 00:03:36 Logon: m_lpMAPISession->Logon returned 80004005
 09/02/08 00:03:36 OPEN DATABASE FAILED: -1018 There is a checksum error on a database page.
 09/02/08 00:03:36 MB2_Chgdir:Logon returned e000fea9
 09/02/08 00:03:36 LP_ENV::MsgError: error 0xe00084af processing object Mailbox Store
This error comes up when the job starts writing data to an IMG File and during backup, the disk is full, so the job tries to delete an expired IMG, and this is when the job fails. If the job has another Storage Group selected, the backup of that storage group will run fine since the previous Storage Group backup has already deleted an IMG to reclaim space.
Upgrade to Backup Exec 12 or later, because the newer versions delete an expired IMG to reclaim disk space prior to the start of the backup, before attempting to write GRT data to a new IMG.