[rdiff-backup-users] Understanding --verify-at-time

[Dominic]

What is the quickest and easiest way to verify the integrity of 
rdiff-backup repositories? I am not here so concerned about comparing 
them with the original data as having assurance that the repositories 
can be used for retrieval of historic data i.e. --verify-at-time.

If --verify-at-time encounters corruption or some other problem, does 
rdiff-backup always return a non-zero error? (I presume so, but I have 
no corrupt archives to check this with!)

If I use -verify-at-time 1Y and this returns without error does this 
confirm that *all* files in all backups in that repository taken 
*during* the last year are valid? Or is it possible that I have a valid 
backup for one year ago but a corrupt backup, in the same repository, 
for six months ago?

I notice that if I have a repository with the first backup say 3 months 
ago and I use -verify-at-time 1Y it still reports 'Every file verified 
successfully', even though there is no backup one year ago. This seems 
strange to me; it presumably means that if all archives more than 3 
months old had been deleted from a backup set, this would not be spotted 
by using --verify-at-time 1Y which would still say 'Every file verified 
successfully'?

Thanks for any advice.

Dominic