Direkt zum Inhalt

Questions about Validating backups

Thread needs solution

My primary backup scheme is to do a Full backup of my entire C: (boot) drive on Sunday and then Incremental backups for the next 6 days of the week. I am still doing .tib backups, not .tibx. Questions:

1. If I ask ATI to validate e.g. the Thursday .tib backup file, it must then also check every incremental-file-and-back-to-the-Sunday-full, correct?

2. Why does Help state this: "We recommend that you validate system partition backups under bootable media. Other backups may be validated in Windows."?

I would like to know why I apparently shouldn't validate my backups, which are of course to another drive, using the Windows application of ATI.

0 Users found this helpful

Tom, you are correct for point 1. validation is pointless unless the whole backup chain is included as a single incremental is useless on its own!

Not sure that I have seen the advice documented for point 2. but to be honest, I do not use validation regularly at all.  I see no point as it does not guarantee that the backup will produce a successful recovery!

Validation is purely a mechanism for confirming that the data written to disk by the backup process has not been changed since written.  It does this by reading each block of the file from disk, then calculating a checksum for that read block and compares the checksum with one that is embedded in the file.  If the checksums match, the file is unchanged from when written to disk.

If the backup was created at a time when any file system corruption was present and not resolved, or when malware was present etc, then it will include the same in the backup image!

When preparing to do a recovery, then it can be helpful to validate the backup chain using the rescue media to check that all is good as far as can be checked, but with the understanding that any latent issues in the backup data may be discovered during the recovery process.

Thanks Steve, I vaguely recall a time with some very early releases of ATI that there were problems with backups and validation became very important, but in at least my case I think the issues were with computer hardware errors i.e. as if USB ports were not up to being reliable for these very very (relatively) large .tib files. It seems to me I recall too using shortest, most robust USB cables and connecting as directly to a PC motherboard as possible so that there was no chance of noise or etc. in the wiring chain.

Regardless of whether #2 is problematic or not, I used Validate again this morning to assure that my incremental chain had not been broken by my screw-up in the other thread here.

;-)