TeraByte Unlimited

I'm seeing an odd but apparently inconsequential (the backup still completes with Error Code 0) message in the log when doing a differential against a 2.97c-made full backup (coincidence or not, this never happened before 2.97c). The batch files involved haven't changed in a long time. I'm only seeing this happen with a differential made against one volume as opposed to another, which also doesn't make sense since they're done in the same way.

The full image was made with /hash, and that creates a .#0 file (Image_Full_Monthly_C_Drive.#0), never a .#1 (the hash files are not nearly large enough to need to go to #1). I have no idea why it would be looking for a #1 when it's found #0.

[8/4/2015 5:15:02 AM] Image for Windows 2.97c
[8/4/2015 5:15:02 AM] Starting ...
imagew.exe /b /uy /um /sch:"F:\IFW_Backups" /base:"F:\IFW_Backups\Image_Full_Monthly_C_Drive" /f:"F:\IFW_Backups\C_Differentials\Image_Diff_C_Drive_20150804" /v
[8/4/2015 5:15:15 AM] PHYLock is waiting for drive writing to stop ...
[8/4/2015 5:16:49 AM] PHYLock using drive C:
[8/4/2015 5:16:50 AM] PHYLock Started
[8/4/2015 5:16:50 AM] PHYLock Using Disk
[8/4/2015 5:16:50 AM] PHYLock version 11
[8/4/2015 5:16:50 AM] Backup: Drive 1 (C:) MBR 1 Partition (02) 26388 MiB HPFS/NTFS
[8/4/2015 5:16:50 AM] To: F:\IFW_Backups\C_Differentials\Image_Diff_C_Drive_20150804.TBI
---->[8/4/2015 5:21:25 AM] Cannot open "F:\IFW_Backups\Image_Full_Monthly_C_Drive.#1".
[8/4/2015 5:21:29 AM] INFO: Total Sectors:54042660 Total Allocated:40563129
[8/4/2015 5:21:29 AM] INFO: 2215845 Sector(s) backed up
[8/4/2015 5:21:29 AM] Validating: MBR 1 Partition (02) 26388 MiB HPFS/NTFS
[8/4/2015 5:21:29 AM] From: F:\IFW_Backups\C_Differentials\Image_Diff_C_Drive_20150804.TBI
[8/4/2015 5:21:36 AM] Operation Completed with Error Code 0
[8/4/2015 5:21:36 AM] Stop

Is it just on that one Full backup image or any Full backup taken of that partition?

If just that one, have you tried recreating the hash file?

I just tried recreating the hash file, and it has the same MD5/SHA1 of the original, so that can't be the problem.

I don't keep previous hash files, but it was not throwing this error last week (even with 2.97c, which it was using for a couple of those differentials) with the previous month's image and hash of the same volume. Looking back through the log that goes back over a year, it never happened before this week.

So that would lead me to think it's something about 2.97c-created images and hashes (last month's full backup was not created with the "c" release since it wasn't out a month ago), except that I'm not getting the error with another partition, so that doesn't really make sense.

The one throwing the error is a system (OS) partition, while the one that's not is two data partitions, but that's a bit of a stretch. Still, it's the only difference that I know.

What version of Windows is it?

It may not matter, but are you using PHYLock or VSS?

This is 2003R2.

Phylock.

Time-wise, can you tell if it's still using the hash file or does it ignore it and take longer?

Do you see the same error if you create a differential based on that backup using the GUI instead of the batch file?

If you create another full backup of the Windows partition and then create a differential based on it does it report the error? Just trying to determine if it's something specific to that particular backup.

It's such a small image (well under 1GB) that it's very difficult to tell by time since there wouldn't be much difference either way.

When I tried it with the GUI though, I actually saw what's happening, happen, right near the end. Maybe you can interpret what it could be. Clicking OK is an infinite loop, and clicking cancel allows the backup to complete (successfully, as always, and with the same log entry as before).
http://i.imgur.com/dvi3M5L.jpg

I will try creating a new full backup soon to see whether there's some freak condition causing this in the one that I have. I could use last month's full backup, but I'd rather create a new one with "c".

Creating a new full backup solved it.

I wonder how something could have been wrong with a full backup that was successfully created?