On Jun 24, 2008, at 1:18 AM, Kendall Shaw wrote:

> I think I can't afford a tape backup system for my home (I think the
> average computer user can't afford a tape backup system sufficient to
> backup their computer, these days).

I bought a number of used DLT7000 tape drives and tested them.  I kept
a few good ones and use them.  Prices varied betwen $50 and $150.

But I fear most of this thread has digressed from your original question
which I think is asking for a strategy given your set of constraints.

> So, I can only afford to keep backups for the purpose of recovering  
> from
> hard drive failure, or restoring a file I deleted within the last week
> or so. I have a "backup computer", that runs bacula, and I think I  
> want
> to have it rotate through volumes on the disk, where backups older  
> than
> a week vanish.
>
> How do you do this?
>
> What I'm doing now is having it do a full backup every week, during
> which it's wiping out the previous week's full backup, since I'm not
> ready to settle on having approximately double the disk space used  
> to be
> protected during that period. I also am not sure yet if I have enough
> disk space to backup the other clients, even without using double the
> disk space.

You really should have one full backup at all times, even when the
new full backup is being created.

Consider not backing up data that can be 'easily' recreated.  For  
example,
the operating system.  I backup only the data that cannot be easily  
recreated,
such as source code, configuration files, etc.  I do not backup the  
operating
system which I can reinstall from CD.

> When I add another client it's going to be close to 24 hours per  
> week of
> backup activity for no purpose. Is there a better way?

Why do you think there is no purpose?

> What I think would be ideal would be to have one and only one full
> backup, and then have rotating differential and incremental backups.
> But, if that means that I have to have file information continuously
> accumulating, my catalog will get too big.

Yes, you could do that, by putting Full, Incr, and Diff backups into  
different
pools, each with different Volume retention periods

> Is there a way to just have file information saved for job #1 (the  
> first
> differential that got bumped up to a full backup) and allow later jobs
> to be autopruned?

Look at File retention versus Job and Volume retention.  These settings
dictate how long the respective data stays in the Catalog.

-- 
Dan Langille -- http://www.langille.org/






-------------------------------------------------------------------------
Check out the new SourceForge.net Marketplace.
It's the best place to buy or sell services for
just about anything Open Source.
http://sourceforge.net/services/buy/index.php
_______________________________________________
Bacula-users mailing list
Bacula-users AT lists.sourceforge DOT net
https://lists.sourceforge.net/lists/listinfo/bacula-users