Welcome to SparkyLinux forums
Zapraszamy również na polsko-języczne Forum

Author Topic: 7 rules of home backups to save me hours next time  (Read 1743 times)

Offline kanliot

  • Jr. Member
  • **
  • Posts: 88
  • Karma: 0
7 rules of home backups to save me hours next time
« on: April 08, 2019, 10:17:32 am »
After using a laptop for a year, I got my desktop out of storage and tried to merge back some of my "important" data.   I thought I was saving disk space by pushing unneeded files off to an external disks.  What I was really doing is wasting time by making a mess!

Rules of backups:

* people say one backup isn't enough.  This is probably true.
* Automated backups will delete data
* You will forget what your backup directories are for, in about a week.  This will leave you thinking endlessly.

    (Try) and Imagine my situation!  I have a directory full of stuff, some of it's things that I've saved, some of it's things I've discarded,   But which is which?
* Label backups with the current date!
* Label backups with the reason you made them!
* Mounted backups will be confused with your current workspace, and you'll carelessly end up losing your workspace data.
* Disks will fail, but you won't lose data, only time!

Btrfs is ok.  The only problem I had was so long ago, it hardly matters.
Probably the best backup strategy is to have two tiers of backup.    The first tier of backup ensures that when you stop on a project, you don't immediately lose that project data due to user error.
The second tier backup covers everything in general.

The second tier would be an automated backup that occurs even if you don't think about it.    So the first tier would use Btrfs snapshot, (not using it), but the second tier would use rsync ( I'd use XFS snapshots, but I don't have hardware raid).  Right now, I have one sciprt that zips everything that I work on into a zip file, then a second tier of script that uses rsync to backup the filesystem.
« Last Edit: April 09, 2019, 07:21:29 am by kanliot »