Re: student question: incremental backups (level 0 and level 1)

From: joel garry <joel-garry_at_home.com>
Date: Wed, 25 Jun 2008 13:15:40 -0700 (PDT)
Message-ID: <b42e4167-710f-4071-98a0-ecbfde41d771@25g2000hsx.googlegroups.com>


On Jun 25, 10:02 am, Holger Baer <baer_at_dont_ask.com> wrote:
> fitzjarr..._at_cox.net schrieb:
>
>
>
>
>
> > Comments embedded.
> > On Jun 25, 4:58 am, pellicleund..._at_hotmail.com (obakesan) wrote:
> >> Hi
>
> >> so far my oracle experience has been on 8i with only PL/SQL development on 9
> >> and later.
>
> >> I'm just going through some questions in preparation for a 10g Admin 1 exam
> >> and find (among others) a question for which the answer seems to be less than
> >> right. Of course I can simply memorize the answer,  but in the interests of
> >> developing knowledge (rather than simply passing exams) I thought I'd seek
> >> some suggestions here.
>
> >> The question poses: You perform differential incremental level 1 backups of
> >> your database on each working day and level 0 backups on Sundays. which two
> >> statements are true about the differential incremental backups.
>
> >> the answer I have issue with is: the backup performed on sundays contains all
> >> the blocks that have everbeen used in the database.
>
> >> I would have thought that it is possible for blocks to be used and then become
> >> free, these blocks should then not need backing up ... no??
>
> > No.  They are below the HWM so they do get backed up.  Any block below
> > the HWM has the potential to contain data.  Thus, all blocks below the
> > HWM are backed up.
>
> >> I'd have thought
> >> it was also possible that eventually (despite being used and freed) this could
> >> eventually encompas every block in a datafile.
>
> > It could.  And the issue with that is?
>
> >> thus the word EVER seems to be making this a wrong answer (which it is not
> >> according to my review software).
>
> > It's wrong only in your mind.  It's correct in terms of the concepts
> > and functionality of RMAN.
>
> >> I'd be happy to learn what I've overlooked on this ... thanks :-)
>
> >> See Ya
> >> (when bandwidth gets better ;-)
>
> >> Chris Eastwood
> >> Photographer, Programmer Motorcyclist and dingbat
> >> blog:http://cjeastwd.blogspot.com/
>
> >> please remove undies for reply
>
> > David Fitzjarrell
>
> The documentation is a little bit clearer (http://download.oracle.com/docs/cd/B19306_01/backup.102/b14191/rcmcon...
>
> <quote>
> Multilevel Incremental Backups
>
> RMAN can create multilevel incremental backups. Each incremental level is denoted by a value of 0 or 1. A level 0 incremental backup, which is the base for
> subsequent incremental backups, copies all blocks containing data.
> The only difference between a level 0 incremental backup and a full backup is that a full backup is never included in an incremental strategy.
>
> </quote>
>
> The  important point to note:
> - a level 0 incremental backup copies all blocks containing data
>
> This does not mean all blocks that *ever* contained data, think of truncate table for example. So the OP is right to express his doubts about the wording,
> even if it's not in the sense he originally meant. Another example: drop a table/index/partition and you're likely to to have lots of blocks that once contained
> data but now don't. They shouldn't (and to the best of my knowledge don't) get backed up.
>
> Cheers
> Holger

Playing a bit of semantics games here, but if a block is no longer under the high water mark, I'm thinking it is not considered ever to have been used... then there's
http://sysdba.wordpress.com/2006/04/28/how-to-adjust-the-high-watermark-in-oracle-10g-alter-table-shrink/

I think the OP is correct, this is an example of the kind of crap questions OCP engenders, specifically antithetical to the kind of thinking a DBA needs to do, someone who actually knows how this works could get it wrong.

jg

--
@home.com is bogus.
Lies.  Damn Lies.  Statistics.  Correlation.
http://www.wired.com/science/discoveries/magazine/16-07/pb_theory
Received on Wed Jun 25 2008 - 15:15:40 CDT

Original text of this message