WHy can't I validate full image plus last differential?

I do weekly full image backups plus daily differentials. I also fully copy my backup disk each week to retain a copy offsite. Before I do this I delete the intervening differentials so that I keep only the initial full image backup as well as the last differential backup so as to save disk space since I keep several weeks of backups on the disk.

I am very concerned that I cannot verify a backup containing only the full image and the last differential backup as without verification I don't feel comfortable that I could restore the image to the state represented by the full image plus the last differential backup.

I can verify the initial image but think it is not verifying the last differential and if I try to verify the last differential it fails saying it wants the previous differential that has been deleted as it should no longer be necessary.

Is there ANY way to verify fullbackup.tib + fullbackup6.tib without the intervening differential files? If not this is something that needs to be fixed by Acronis as if I can't verify why should I feel I can restore fullbackup.tib + fullbackup6.tib If you can't do this then differentials are really not worth much as you have to have them all and that is a LOT more space than incrementals. PLEASE FIX THIS ACRONIS!!!

Comments? Experiences? Thanks, Dave

 

TI has been this way for a long time. It treats differential backups the same as incremental backups. They all have to be there to Validate.

This "bug" has been requested many times to be removed, but Acronis has left it.

You should be able to mount/explore the image, but the only way to test if it's completely intact would be to restore it.

 

They're NOT the same so should not be treated the same. What is the point of differentials otherwise. It would take me about 10 minutes to write code that recognized there were other files of a similar name in a directory even if missing the intervening files.

Acronis support - what do you have to say about this. What good are differentials if you cannot verify them. WHy should I feel like I could do a restore if I can't do a verify?

Please comment and justify or fix this anomoly.

Thanks,
Dave

 

Also there should be consistency between validate and restore. They should both work or not work in the same fashion. And if restore does not work on backup1 + backup6 then differentials are totally worthless. So which is it? Do restores work? IF so then validate should as well. A bug pure and simple and extremely easy to fix. If all the files are there then validate them all else validate what is there.

 

Acronis support - would you please address this? Thanks, Dave

 

Hello Acronis Support....

Why can't I validate what I can supposedly restore? Or can I actually restore a full image plus the last differential. I suppose I need all the differentials to restore also huh

 

As said in Post #2, this is just how TI is. Acronis has not seen fit to change it even though many, many people have complained about it.

You don't need the "inbetween" differentials in order to do a restore or mount/explore. You just need the FULL and the last DIFFERENTIAL backup. However, if you don't have the "inbetween" differentials, the only way to know the backup images are valid is to restore them since you can't Validate them. Some would say that this really doesn't matter since a successful Validation doesn't automatically mean a successful restore, but personally, I perfer to Validate.

 

Yes I understood you but am not happy about the situation and want to complain and am.

It is poor design to expect that I can restore what I can't validate. If one is complacent about this why should they fix it?

I want them to fix an obvious problem in the design of differential backups.

If you don't like this behavior please complain to Acronis.

In any event I want Acronis to at least explain their reasoning and TRY to justify why they go thru all the trouble of supporting differentials and then make them unusable.

Dave

 

The differentials are not unusable. You can use them to restore. Validation is the issue if you delete one of your differentials.

 

As I said - I understand that I am supposed to believe that. Of course why have differentials if you have to keep them all in order to validate that the last one works? Or that a "copy" of the incrementals is valid before you archive it off site? They take up a LOT more space than incrementals. If you can supposedly restore, which by the way I believe is probably true, then why can't I validate? And why should I have to keep all differntials just to take another one? If I want to cleanup folders in respect to disk space so that I can maximize the backup time span then I should save the last full image plus the last differential and archive the pair and after doing so I want to validate it. Guess you never had a bad network card huh? Try the Intel 1G Pro sometime. Sorry not, but I want to validate what I am expected to be able to restore. And I would like Acronis to address this issue. It is a bug or at minimum very poor design that in my opinion that can EASILY be fixed and have done software development for many years so know this would be a very minor change for them.

 

I understand you would like Acronis to fix this issue. But continuing to post it here will most likely not have the desired effect. Posting here may https://www.wilderssecurity.com/showthread.php?t=85595 help or opening a support issue with them.
And no I've never had a bad Network card whatever that has to do with the issue.

 

I have in fact opened a support issue several days ago and have not heard from Acronis. If enough folks are aware of this issue and complain then perhaps Acronis will fix it. And what a network card has to do with the issue is that after copying a set of backups to another drive I want to verify the copy was OK since I had an issue with an Intel 1G Pro card corrupting data that made the backup unusable. Problem was fixed going to a Netgear 1G card. I don't want to copy all the differentials as again that is not the point of differentials. Have an open mind and realize this is a problem. One should not be expected to be able to restore what one cannot validate. I could tell you several other scenarios that could go wrong but bottom line is I want to validate what I might have to restore.

 

You could use an MD5 checksum program on the source and the copy and make sure they match. If they do, then the copy should be okay.

You could also run them through a binary file compare ( fc /b file1.ext file2.txt ) and see if they match.

 

Yes that certainly works in a copy scenario and is how I proved that the Intel enet card was corrupting data sometimes on large 30+ gig files. Acronis validation is what originally alerted me to the fact that my Intel enet card was corrupting data since it would often not verify after the backup. I ran exhaustive memory and hard disk tests and only replacing the enet adapter fixed the problem which allowed Acronis to validate from the network share consistently. Another way to validate the backup files is to use Acronis which is the means that I prefer in my new backup environment since that is built into the Acronis backup job but unfortunately does not work if I have already removed all intervening and no longer needed differentials other than the last. Why keep them as I can't fit as many backups on the disk if I do. I always verify as the last step of a backup. If I have one good differential there is no reason to keep the others than the last other than this problem with Acronis which could easily and should be fixed by them. Don't see any reasons to make excuses for them. WHen I write software and someone points out a bug or design defect I am thankful and fix it as soon as I can depending on urgency and priorities. But at least I acknowledge it right off even if I don;t have the time to fix it immediately. Would like them to do the same and schedule it into their plans which makes their system a lot more usable. Dave