[svlug] Backup strategies (was: Reminder: SVLUG mailing list server failure risk, gap in backups)

[Scott DuBois]

On Wed, Dec 31, 2014 at 12:13:01AM -0800, Rick Moen wrote:
> Hey, folks, I know that topic drift happens, but might I suggest that people
> change the Subject header if they digress from dire threats to SVLUG's
> mailing list server to, e.g., chatting about tech. groups in Orlando and El
> Paso?   Maybe even a new thread?
> 

Yeah, I notice this a lot in all kinds of discussion forums.

<raises hand> Guilty myself on more than one occasion.

> If you're worried about missing something important, the 'safe' 
> option seems to be the 'grab everything' one.  And that, in turn, 
> motivates people to attempt pointless things like backing up 
> /proc, /sys, and /dev.

This process seems to be 'fairly' common by a lot of people under the
concept that writing the script, or running the command, to just grab
'everything' is much easier than surgically removing the relevant parts.
The concept is betting that the odds of needing the backup is remote,
and in such an occasion as it would be required, dissect the backup of
the necessary components to rebuild.

Seems more like a 'reverse' logic to me as the time crunch is when the
backup needs to be restored vs. making the backup.

> (Same for /run, which I gather must be an Ubuntu barbarism?  Should still be
> /var/run, IMO.  Wasn't broken, shouldn't have been fixed.)

I'm noticing such variations more and more. Debian stable is beginning
to look more interesting by the day.