Re: deduplicating file systems: VDO with Debian?
On Wed, 2022-11-09 at 11:37 +0100, didier gaumet wrote:
>
> I am no expert (in Linux, backporting or anything else) and cannot emit
> a viable advice about what your backup plan should be. You are in better
> position to evaluate your needs, your means and design a satisfying
> backup plan accordingly.
>
> What I was underlyning is that in my opinion you are confusing
> deduplicating during backup and incremental/differential backups.
> (Perhaps in your context that has no consequences and is thus unimportant).
> To *me* what you are talking about is incremental/differential backups,
> not deduplicating backups.
I don't know why you think that. To clarify, I haven't been making incremental
backups. Instead, I keep two full backups that were created at different times.
Making incremental backups through snapshots or other means is very different.
Perhaps it's a better solution because it needs less disk space.
> For example, I am myself using Deja-Dup (based upon Duplicity, itself
> based upon librsync) for my basic home laptop backup: it's incremental
> backups but I would not call it deduplicating backups.
>
> The Wikipedia deduplicating paragrah of their backup article has an
> example of 100 identical workstations having a backup storage need
> divided by 100 by deduplication.
>
> Wikipedia "deduplicating" parapragh of their backup article:
> https://en.wikipedia.org/wiki/Backup#Deduplication
> Wikipedia "incremental" parapragh of their backup article:
> https://en.wikipedia.org/wiki/Backup#Incremental
> Wikipedia "differential" parapragh of their backup article:
> https://en.wikipedia.org/wiki/Backup#Differential
I've made it simple and haven't done any of this :) It's only two full copies
that get updated interchangeably.
This isn't to say that it would be better or worse. Only when you have two
copies of almost the same data, you can easily think that it would make sense to
deduplicate them. What's confusing about it?
Reply to: