Re: [LUG] Borked system rescue plan

Not sure if I've understood correctly what you want, but I'm sitting here now using Clonezilla to image my root partition, before one of my 4-5 times yearly Arch (pacman -Syu) updates. It has been a very reliable & straightforward tool, in my experience.
Whilst Linux systems tend to be recoverable, sometimes it takes a whole lot of Googling, chrooting & patience.
You can save yourself a lot of time, sometimes, by having a current enough snapshot.
Of course, then you don't learn as much (fixing things).

Daniel Robinson <manipula@xxxxxxxxx> wrote:

As we learnt last night, Jessie will be stable in a few months time and so the particular application I want to use (Steam) will be working on a Debian stable release.

Things break when trying to update glibc and that sort of thing and my knowledge of Linux is not good enough to recover the OS when this sort of thing goes wrong.

Again, This is why I find snapshots the bast way to recover from such a disaster!

On 21 October 2013 12:55, Gordon Henderson <gordon+lug@xxxxxxxxxx> wrote:

On Mon, 21 Oct 2013, Daniel Robinson wrote:

Honestly, I'm a noobie using testing and experimental repositories, I break
it all of the time. Having a snapshot that takes 30mins to build for me in
my working environment is by far the quickest and safest way (for me).

I have kind of answered my own question by using something called redo.
http://redobackup.org/

Never heard of it Âuntil now.

It is perfect for my needs, I was hoping that somebody here does backups in
this manner and would point me in the direction of something like this.
(Again I was wrong)

You need to work out why you need a backup and what you're actually backing up.

Using experimental distros, installing packages that "break" - well, you're in a tiny minority there (I reckon).

And since you're using Debian, I still don't see why you need to re-install when apt-get/aptitude/dpkg are supposed to take care if it all for you - if you install a new package, and it doesn't work, then just un-install it...

In my 20 years of using Linux I've not once destroyed an installation to the extent that I needed to restore the OS backup. I've had servers fail, but then I've done an install from scratch as part of the repair process. I also usually do an install from scratch when I get new hardware. (e.g. to upgrade my desktop)

But then I've always used Debian Stable since it was ... stable.

I do find commands and scripting still a bit frightening when it comes to
back up and restore but now have a recovery plan I feel comfortable and
accustomed to.

This now enables me to explore the other back up options with a safety net
in place just in case things go horribly wrong.

But thank you Gordon for taking the time and effort to reply to my topic, I
will now explore the options you have suggested. Thank you

Why not create virtual servers to test your experimental installs? Then it would be a simple matter of copying the filesystem tree (if it's at the filesystem level outside the VPS), or the single file that contains the VPS filesystem to make a backup.

Gordon

--
The Mailing List for the Devon & Cornwall LUG
http://mailman.dclug.org.uk/listinfo/list
FAQ: http://www.dcglug.org.uk/listfaq
-- 
The Mailing List for the Devon & Cornwall LUG
http://mailman.dclug.org.uk/listinfo/list
FAQ: http://www.dcglug.org.uk/listfaq