Slow recovery of file (one generation back)
I'm running ATI Cloud (6571) on Windows 10. My backups are on an external USB drive.
I want to restore a particular file. If I restore the most current version, it takes a minute or two.
If I try to restore the same file from one generation back (the day before), it takes -- well I don't know how long it takes, because it hasn't finished. It's been half an hour, at least. [Ignore: It still says "Calculating size."] After hitting "Restore now," ATI goes into a trance.
This backup chain is 32 incrementals long, which I admit is pretty long; but doesn't ATI have a database that tells it which tib file to look in? It should know that the previous day's version is in the previous tib file (#31), so shouldn't this be pretty quick?
Is it going through all of the previous tibs in the chain?
Is this ever going to finish?


- Log in to post comments

Actually, my original post was incorrect. I had tried this several times, and gotten myself confused.
I hit the "Next" button and then the "Restore now" button. That's when ATI went into a trance.
I'm not sure why any kind of snapshot would be involved in restoring a file. I certainly don't understand why there wouldn't be any problem restoring the most recent version of a file, but restoring an older version would fail. I would think that restoring an older version of a file would be a good reason for doing backups in the first place.
I know that Acronis dropped a lot of features to make the product more user friendly, but I wouldn't think that restoring files would be one of them. I'm extremely tech savvy, but turning VSS on and off is not something I do on a regular basis. If their target market is the less sophisticated user, turning VSS on and off would seem to be a strange way of making the product easier to use.
If people aren't able to restore files, what good is a backup? I'd think Acronis would be interested in fixing this ASAP.
This whole thing is bizarre.
- Log in to post comments

Jerry, I agree with your comments for this issue and also don't understand why VSS should be involved in a fairly simple recovery / restore operation but the other user post I referenced would seem to suggest that VSS is involved.
I am sure that Acronis has not dropped the ability to restore files - there are plenty of users in these forums who report successfully doing just that albeit they are coming to the forum to ask about other aspects of the product etc.
Have you tried just opening your specific incremental backup archive file in Windows Explorer and accessing the specific file / folder contents via that method?
If you want to persue this issue more directly with Acronis, then you would either need to open a support ticket with them or alternatively, use the Feedback tool in the ATIH GUI Help section along with an Acronis System Report.
- Log in to post comments

Well, on a hunch I checked to see if VSS was running, and it was not. I started the VSS service, and tried to restore that file again. It worked like a champ.
This is the exactly opposite to the description in 115636: Won't recover, calculating size
VSS is used when restoring files, because it handles various possible problems: file in use, for example. It is a manually started service, so if ATI wants it, ATI should turn it on. That didn't seem to happen.
I hadn't tried any of the other possible ways to recover the file. I wasn't really worried, since the file is backed up to the cloud. I just tried the "natural" way and got stubborn when it didn't work.
Since I have subscription support, I'll contact them when I have a chance.
- Log in to post comments

Nuts!
I tried to duplicate my results. I started VSS, and then I launched ATI and tried to do a restore. It got stuck just the way it had before. It looks like VSS is a red herring.
This seems to be a timing problem within ATI. If I select a recovery point date and wait for the file list to reflect the changes (file date, whatever), everything works as expected. If I select a file before ATI has retrieved the data for that recovery point, it will hang. Even though ATI says it is calculating, you can kick it free by unselecting the file(s) or selecting a different recovery point.
If you can't spot anything changing in the file list, waiting 10–15 seconds before selecting files seems to be long enough.
- Log in to post comments

As far as I know, Acronis has to parse through the entire Incremental backup scheme to compare the differences in order to find the version of the file and where it exists. For instance, if you changed the file in incremental 2 and it hasn't changed since then, it still needs to check all of them to make sure there is not a newer version in say version 32. Incrementals for life is a bad idea for these reason, so I'd recommend setting a full to occur at some point to reducd the number of incrementals it has to search through.
That said, what happens if you try to mount the backup file (assuming it's a disk/parition backup) and then copy and paste the file out of it that way instead of using the Acronis GUI interface? Or, what happens if you double click on the backup.tib file and let it try to open up in Windows File explorer to copy and paste from insted of using the Acronis GUI interface?
- Log in to post comments

It is a disk backup, full plus incrementals.
If I try to mount one of the incremental tibs (right click, TrueImage, Mount), the ATI mount wizard complains that it can't assign a drive letter to a partition from the backup archive.
If I open an incremental tib using Explorer, I can see all of the backups (up to that point). If I open one of the backups, I can see something labeled C:; but either it's empty, or Explorer can't see what's inside.
I can mount, or open, the full backup. That doesn't get me where I want to go, since I wanted to get a version from an incremental.
It looks like waiting for the ATI GUI to finish parsing the recovery point is the only way I can get what I want. That brings me back to this being a bug in the GUI. If a GUI is busy doing something, and can't deal with any user input, it should disable the troublesome bits. It shouldn't hang if the user jumps the gun.
- Log in to post comments