Page 1 of 1
Continue on error?
Posted: Sun Nov 05, 2017 5:18 pm
by n64
Can an option be added for continue on error? Or even specifically for when files no longer exist?
A couple times a week the backup gets this error
Code: Select all
Can't open "\??\C:\Users\user\AppData\Roaming\Mozilla\Firefox\Profiles\vkwnivsk.default\storage\permanent\chrome\idb\26516546a5sdkjlkasm4o0tc.h.sqlite-shm" for reading (status=c0000034): Object Name not found
[WARNING] Truncating "W:\Backup 44.wim" to its original size (382657270 bytes)
ERROR: Exiting with error code 47:
Failed to open a file.
and it just aborts the backup, which seems dumb to me. Maybe abort if it thinks it's going to corrupt the backup, but why would you abort just because it can no longer find a file?
Anyways yes I know I could use the --snapshot, but I rather avoid that and requiring admin rights.
Re: Continue on error?
Posted: Sat Nov 11, 2017 7:10 pm
by chungy
wimcapture isn't really great to run against trees that are in use and being modified. On Windows, you can add --snapshot so that the OS exposes a point-in-time view of the tree that shouldn't change when software modifies anything.
Re: Continue on error?
Posted: Sun Nov 12, 2017 6:33 pm
by n64
I mentioned not wanting to use --snapshot because it requires admin rights.
I just feel it's a bit weird it fails because the file is missing, not locked, but missing and then it completely aborts the backup. Any other backup application I have used would just continue on error and report it at the end.
Re: Continue on error?
Posted: Mon Sep 18, 2023 2:05 pm
by BiatuAutMiahn
Apologizes for the necro bump. When attempting to archive a user's folder with snapshots enabled there are files that get locked due to both OneDrive, and a filesystem write filter not being available. The result is most of the user's files being read-only and access denied. I need to be able to ignore all of these errors so I can recover the user's data that isn't locked.
Currently I use a script that generates an exclusion list by: Gathering a file list, excluding symlinks, and various space hogs that aren't important for the user like OneDrive, DriveFS, etc. The problem is that this is very slow, and further more to then check each file for read access, which will most likely fail then a process has locked the file. I'd rather not have to implement VSS in my script to avoid these errors.
Re: Continue on error?
Posted: Wed Sep 20, 2023 8:58 pm
by synchronicity
Reparse points are supposed to be captured as-is, not dereferenced. Any idea why that isn't working for your case?
Re: Continue on error?
Posted: Thu Sep 21, 2023 4:03 pm
by JFX
Sounds like the problem with OneDrives cloud files.
Wimlib tries to read a placeholder file and the cloud provider failes as the shapshot of
the volume is readonly.
Personally I exclude all files with FILE_ATTRIBUTE_RECALL_ON_DATA_ACCESS attribute for capture operation.
If one detaches the CldFlt filter from the volume, than wimgapi can capture a OneDrive folder.
But wimlib would ignore the content of the folder and just capture the root folder (unknown reparse point).