On Thu, Oct 25, 2012 at 10:38 AM, Timothy J Massey <tmas...@obscorp.com> wrote:
>
> Also, keep in mind that the original question shows a fundamental lack of
> understanding in the way BackupPC works. BackupPC is *NOT* a simple
> replacement for a tape drive. Swapping the entire pool weekly will undermine
> the way BackupPC works: it's not designed for that. And using preconceived
> ideas of traditional backup to shape the way BacukpPC works is not a path
> that leads to success.
>
As far as backuppc goes, it should work fine if you simply swap the
whole archive. It will just catch up in the same way it would if you
had simply shut the server down for the duration of time that the
swapped-in disk had not been present. However, you need to
understand the tradeoff that since everything is not redundant, a disk
failure will lose the data not present on the other drive(s) and you
may not always have the thing you want to restore available. You'd
need to stop the backuppc server and unmount the partition before
removing it. And if I were doing it I would probably rotate at least
3 drives to be sure they were never all in the same place at the same
time. If you swap at the end of a week, letting everything catch up
over the weekend you'd have a pretty good chance of always have a
recent copy handy.
> 1) Constantly mirroring the pool and occasionally breaking the mirror to take
> it off-site. (Option 1 above). A variation of this would be to take the
> pool down and make a copy of it, then bring it back up (or use LVM snapshots
> to reduce the downtime). This variation is left as an exercise for the
> reader: there are a *lot* of unexpected details in that answer: problems
> with file-level copy will most likely require block-level copies, LVM
> snapshots present performance and reliability issues, etc.
This works, but has the down side of taking most of a day to re-sync
the replaced mirror with the drives being too busy to be useful for
much else during that time.
> 3) Have two BackupPC systems that *both* back up the same hosts in parallel.
> A variation of this would have the BackupPC servers in different physical
> locations. This variation just about requires the use of rsync, and even
> then is not always practical (if the data changes per day/week/whatever are
> too vast, the bandwidth between them too limited, or there is simply too much
> data to be able to back it all up twice in a reasonable amount of time).
This is really the best solution if you have a host in another
location and enough bandwidth to make it feasible.
--
Les Mikesell
lesmikes...@gmail.com
------------------------------------------------------------------------------
Everyone hates slow websites. So do we.
Make your web apps faster with AppDynamics
Download AppDynamics Lite for free today:
http://p.sf.net/sfu/appdyn_sfd2d_oct
_______________________________________________
BackupPC-users mailing list
BackupPC-users@lists.sourceforge.net
List: https://lists.sourceforge.net/lists/listinfo/backuppc-users
Wiki: http://backuppc.wiki.sourceforge.net
Project: http://backuppc.sourceforge.net/
