Bacula-users

Re: [Bacula-users] Bacula and deduplication

2010-03-02 13:24:54
Subject: Re: [Bacula-users] Bacula and deduplication
From: Steve Polyack <korvus AT comcast DOT net>
To: Beck J Mr <james.beck AT shunsley.eril DOT net>
Date: Tue, 02 Mar 2010 13:06:55 -0500
On 03/02/10 11:50, Beck J Mr wrote:
> I think it's only available in V5.0.x. Even then, it's in it's infancy and, 
> as far as I can tell, doesn't work like all other deduplication methods I 
> have seen; where all files that have the same hash value are deduplicated. I 
> believe this is not the case with Bacula's deduplication.
>
> As far as I understand it, deduplication only occurs on files from the same 
> client as the "Base Job". Then, if a file is altered in any way (file 
> permission, moved to another folder etc) deduplication isn't effective and 
> the file is backed up again.
>
> Also, the Bacula docs say that using base jobs currently makes restoration 
> very difficult!
>
> Please, somebody correct me if I am wrong here. Dedup is something I am 
> interested in if it works in the right way.
>    

I believe you are correct.  I posed some of the same questions and did 
not receive an answer.  As far as I understand, the flags/attributes 
which will achieve file deduplication are configurable (and share the 
same characters/meaning as Verify/Accurate backup jobs), but the file 
path is a hidden requirement.  So, yes, if you move the file to another 
folder which does not exactly match the path in the Base job, it will 
not be deduplicated.  Permissions, however, can be ignored.

The options for which Bacula will consider a file for deduplication are 
the same listed under verify=<options> in the fileset resource 
documentation here: 
http://www.bacula.org/en/dev-manual/Configuring_Director.html#SECTION001470000000000000000

If anyone who is familiar with how this stuff works on the low level, 
I'd be particularly interested on a solid answer on whether the path 
matching with the Base job can be disabled or worked around.

> James
>
> -----Original Message-----
> From: Carlo Filippetto [mailto:carlo.filippetto AT gmail DOT com]
> Sent: 02 March 2010 15:45
> To: bacula-users AT lists.sourceforge DOT net
> Subject: Re: [Bacula-users] Bacula and deduplication
>
> very verbose
> :)
>
> This feature is available only with the last version?
>
> I have the 2.4 and 3.0.3
>
> Thank's
>
> ---
>
> 2010/3/2 Marc Schiffbauer<marc AT schiffbauer DOT net>:
>    
>> * Carlo Filippetto schrieb am 02.03.10 um 14:27 Uhr:
>>      
>>> Hi,
>>> bacula have a feature for the deduplication?
>>>        
>> yes.
>>
>>
>> --
>> 8AAC 5F46 83B4 DB70 8317  3723 296C 6CCA 35A6 4134
>>      
>
>    



------------------------------------------------------------------------------
Download Intel&#174; Parallel Studio Eval
Try the new software tools for yourself. Speed compiling, find bugs
proactively, and fine-tune applications for parallel performance.
See why Intel Parallel Studio got high marks during beta.
http://p.sf.net/sfu/intel-sw-dev
_______________________________________________
Bacula-users mailing list
Bacula-users AT lists.sourceforge DOT net
https://lists.sourceforge.net/lists/listinfo/bacula-users

<Prev in Thread] Current Thread [Next in Thread>