I cam carring out evaluation and test restores of SBS2003 Exchange Server images to dissimular hardware. One of the tests the server fails to boot.

Note: As stated above verify of the image reports OK:

Althought the image itself is good, occasionall a customer's original file system (which is backed up within the image) is corrupt before the backup occurs.  This can cause some odd effects, such as the inability to find NTFS.SYS as you're seeing.

Please mount the image as a writeable volume and then use chkdsk.exe to analyze the mounted image volume and report results here. 

Looks like you hit the nail on the head - chkdsk reports an error on the mounted volume - Unable to run chkdsk /r on the mounted volume because the volume is in use and chkdsk will force a dismount - the disk will then cease to exist.

Looks like I will have to run chkdsk on the physcial os drive. Then take a new image and carry out a test restore.

Thank for the help.
qbit

When chkdsk.exe forcefully dismounts a volume, the volume doesn't "cease to exist."  All that happens in this case is that chkdsk.exe invalidates any currently-open handles which are in use on that volume so that chkdsk.exe can get a full lock on the entire volume.  The volume itself doesn't go away.  It's particularly safe for you to do this type of operation against the writeable mount of a backup image file.  I suggest you take these steps:

1) Mount your backup image as a writeable volume (you have to uncheck the checkbox in the mount wizard that says (by default) that the mounted image's volume will be read-only).  In this case I'm going to assume that you mounted it to the drive letter X:

2) Open cmd.exe and run the following command:  chkdsk.exe X: /F /X

3) When chkdsk.exe has finished repairing X:, go back into ShadowProtect and dismount the mounted image and be sure, within the dismount wizard, that you specify that you wish to save the changes that were made to the volume as a new incremental, and specify a unique name for this incremental (maybe something like C_VOL_CHKDSK_PERFORMED or something like that).

4) Now restore (using HIR) this newly created incremental to your machine.  This new incremental contains the fixes made by chkdsk.  When doing the restore, do NOT check either of the two HIR Advanced options check boxes (don't check "clear configuration" and don't check the one about network conf for domain controllers).  We've learned through experience that these two options cause more problems than the issues they resolve, and we've removed them in the latest 3.1.0.3 recovery environment.  Instead, follow these instructions after the HIR restore:

After finishing an HIR restore of a domain controller, the first boot after the restore should always be done using the Active Directory Restore Mode.  Here are instructions for starting your domain controller in "Directory Services Restore Mode."   When you are booted into Active Directory Restore Mode, you should then follow the procedures outlined in this white paper: 

http://forum.storagecraft.com/Community/files/folders/156/download.aspx 

 
Additional useful links are found int his white paper: 

http://forum.storagecraft.com/Community/files/folders/314/download.aspx