Today's article is about the *minimum* you should do in order to have a proper backup strategy.

[dobeshow]

I have around 10 systems to play with and to back them up i use a dedicated machine (specs don't matter, it's only on when doing a backup) which copies important directories (like /etc /home /var) with rsync. It does either a complete backup or only copies changed files into a new directory and hard links the files which didnt't change since the last backup. It's here if you wanna have a look: backup with rsync
Most settings are hardcoded, but it can be easily adapted.

The stuff i really don't wanna lose is off site on optical media.

Regards,
cs

 

[Anawilliam850]

Is a really good article

 

[KGIII]

I believe you are talking more about engineering redundancy (3 components have to fail before the system fails), rather than a

Recovery Strategy

... which I feel is much more important.

Well, to be fair, the article is about *backups*. How you recover is up to you and depends on the software you use. I wasn't gonna open that kettle of fish. This article is all about the how and why you wanna use the 3-2-1 backup plan (at a minimum). Heck, just sticking an extra copy of your backup in the garage might be enough to recover from fire, theft, and fat fingers.

I, for one, would still like to have my pictures and legal documents (and my ~/) if the house burns down.

That does make me think the site could use a well-written page about Timeshift. *hint* *hint*

 

[dobeshow]

I once did a variation of the infamous

rm -rf . /

...

took me a while to restore it remotely

 

[KGIII]

rm -rf . /

I just ran that the other day. It was in a VM. I wasn't sure, I'd thought Ubuntu had done something that prevented you from doing so, at least not with additional confirmation.

Nope. It happily ran and stopped the system with a cryptic message.

It was just a VM, so I didn't lose anything of value. I probably should have made a snapshot first, but it's not like making a new VM is all that challenging. There was no significant data to lose.

 

[dobeshow]

I just ran that the other day. It was in a VM. I wasn't sure, I'd thought Ubuntu had done something that prevented you from doing so, at least not with additional confirmation.

Nope. It happily ran and stopped the system with a cryptic message.

It was just a VM, so I didn't lose anything of value. I probably should have made a snapshot first, but it's not like making a new VM is all that challenging. There was no significant data to lose.

I'm using a highly customized zsh and the alias i set for rm only included the switch to not cross file system boundries.
I pulled the rug from under a few docker containers. The were not happy about that.. But at least i found out that my backup concept works.
Anyway, there are infinite ways of shooting yourself in the foot and even if you wear shoes with steel caps, the bullet ricochets and hits something else...

 

[KGIII]

But at least i found out that my backup concept works.

See? That's a positive thing. An untested backup is not a backup!