VMDK disk corrupt after discarding snapshots

[Martin Stjernholm]

I had a working windows vista guest where I had made two successive snapshots. After discarding the snapshots, vista won't boot anymore, halting in the Windows Boot Manager with:

Status: 0xc00001 Info: An unexpected error has occurred.

I've not yet been able to verify the integrity of the ntfs file system.

There is more than 25 Gb of free space on the hosting file system. VBoxManage showhdinfo on the corrupt(?) disk image shows this:

UUID:                 be4813bf-f885-4c0d-c49b-c848c3a7039a
Accessible:           yes
Description:          
Logical size:         32768 MBytes
Current size on disk: 22170 MBytes
Type:                 normal (base)
Storage format:       VMDK
In use by VMs:        miffot (UUID: d7316699-797f-467b-b08b-24ef2687b1c4)
Location:             /home/mast/more/vms/windows-vista-x86_64/sysdisk.vmdk

Two logs attached: One before the snapshots were discarded, and one from a broken boot afterwards.

[Martin Stjernholm]

Log before problem

[Martin Stjernholm]

Log after the problem

[Martin Stjernholm]

Log after the problem

[Martin Stjernholm]

Please ignore the last log. It was a duplicate attachment.

[Martin Stjernholm]

I did some things before the snapshot discardings that might be relevant:

1. Took a backup (using cp -r) of the vm directory containing the vmdk and the snapshots.
2. Ran virtualbox 3.0.6 b1 for a short while, but without making any changes to the snapshots or the vm configuration parameters.
3. Removed the original vm directory and put the backup in its place.

Note that I did not backup and restore the .VirtualBox config dir during this. Is it possible that an inconsistency between the config xml file and the vm directory could produce this fatal result?