[prev in list] [next in list] [prev in thread] [next in thread] 

List:       bacula-users
Subject:    Re: [Bacula-users] Tapes to archive
From:       Jesper Krogh <jesper () krogh ! cc>
Date:       2008-03-29 9:04:54
Message-ID: 47EE0636.7050801 () krogh ! cc
[Download RAW message or body]

Jesper Krogh wrote:
> Jesper Krogh wrote:
>> Arno Lehmann wrote:
>>> Hi,
>>>
>>> 22.03.2008 10:21, Jesper Krogh wrote:
>>>> Arno Lehmann wrote:
>>>>>> What I need is a set of job-runs that go to the archive pool, those jobs 
>>>>>> should never be used for subsequently incremental or differential jobs, 
>>>>>> since it is hard to actually retrieve the tapes from the "archive" when 
>>>>>> first shipped there.
>>>>>>
>>>>>> Can I send a single jobs to 2 pools in the same run? That could 
>>>>>> basically solve my problem.
>>>>> ... as this is currently not possible, one work-around is to run one 
>>>>> extra full job to the archive pool. If you really don't want to or 
>>>>> can't use use that pool, you can schedule like this:
>>>>> 1st weekend of month - archive job full
>>>>>                         on-site job differential
>>>> Will this differential job, not be based on the archive-job-full just 
>>>> above and not the earlier diffferential or on-site full?
>>> As long as it's a different job - in my example, on-site vs. archive - 
>>> it won't be based on the archive job.
>>>
>>>> In my case the filesets are quite large.. and does take > 24 hours to 
>>>> run if run as a full-run. (it is another task to speed this up).
>>> Well, that makes this a bit inconvenient, and...
>>>
>>>>> 2nd, 3rd, 5th WEoM   - on-site job differential
>>>>> 4th WEoM             - on-site job full
>>>>> weekdays             - on-site job incremental or diferential
>>>>>
>>>>> That results in a reasonable backup load and restore effort in many cases.
>>>> This is just scheduling two full runs per month, which in general is 
>>>> acceptable. But I have hard arguing that it is acceptable that I have 
>>>> some restores that I cant perform correctly without access to the 
>>>> archive tapes. (Those changes done inbetween the archive-job run and the 
>>>> earlier full/incremental run) They will only be available on the archive 
>>>> tapes.
>>> (That's the reason to set up two distinct jobs.)
>> Ah, now I read the documentation correctly..  it is the "jobname" it 
>> matches on that I need to make different. Thanks alot, that solve my 
>> main problem..
> 
> Can I somehow somehow query the database to confirm this behaviour?
> 
>   14101  Incr        208    48.91 M  OK       27-Mar-08 22:48 Lambda_Daily
>   14113  Full    404,157    7.333 G  OK       28-Mar-08 18:23 Lambda_Archive
>   14122  Incr        221    71.04 M  OK       28-Mar-08 22:54 Lambda_Daily
> 
> 
> I'd like to confirm that the jobid = 14122 has used 14101 as base for 
> the incremental and not 14113 ?

Ok. investigating the catalog I have this file on the Incremental Tape
(Which job 14101 and 14122 spools to)
-rw-r-----     syslog        adm         7755 Fri Mar 28 08:05:11 2008 
lambda-fd:/var/log/mail.log

And if 14113 had been used as a base.. I should have any files on the
incremental tape with dates in the period from Mar 27 22:48 -> Mar 28 
11:23..

So everything is fine.. I works as advertised (as usually).

Jesper
(I just like to check, double check and triple check, especially when it 
comes to backup)
-- 
Jesper Krogh

-------------------------------------------------------------------------
Check out the new SourceForge.net Marketplace.
It's the best place to buy or sell services for
just about anything Open Source.
http://ad.doubleclick.net/clk;164216239;13503038;w?http://sf.net/marketplace
_______________________________________________
Bacula-users mailing list
Bacula-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bacula-users
[prev in list] [next in list] [prev in thread] [next in thread]