Re: size of Incremental Backups verses Redo Generated
- From: kyle Hailey <kylelf@xxxxxxxxx>
- To: ORACLE-L <Oracle-L@xxxxxxxxxxxxx>
- Date: Thu, 21 Oct 2010 13:09:08 -0700
As a follow up, is there a way to use Block Change Tracking to see the
number of blocks changed and thus to calculate the size of the incremental
backup and calculate the relationship between the incremental backup size
and redo generated? Doesn't look like it though from Alex Gorbachev's
excellent paper
http://www.pythian.com/documents/Pythian-oracle-block-change.pdf
there is an upper side formula using the BCT info:
SELECT count(distinct bno) * 32
FROM x$krcbit b
WHERE b.fno = 7 AND b.vercnt >=
(SELECT MIN(ver) FROM
(SELECT curr_vercnt ver, curr_highscn high, curr_lowscn low
FROM x$krcfde WHERE fno = 7
UNION ALL
SELECT vercnt ver, high, low
FROM x$krcfbh WHERE fno = 7)
WHERE (SELECT MAX(bd.checkpoint_change#)
FROM v$backup_datafile bd
WHERE bd.file# = 7
AND bd.incremental_level <= 1) between low and high);
Apparently though BCT tracks 32K chunks that have a change in them but if
the database blocks are 8K, then each 32K will have 4 datablocks and of
those 4, we can’t determine how many have been change – at least one and no
more than 4 so the script in the worst case could be off by a factor of 4.
Thanks
Kyle
http://db-optimizer.blogspot.com/
On Thu, Oct 21, 2010 at 12:45 PM, kyle Hailey <kylelf@xxxxxxxxx> wrote:
>
> What relationship do people see between the size of Incremental Backups
> and the size of the Redo generated between those backups?
>
> Of course it completely depends on the system. If the system is hitting the
> same blocks over and over again then the redo could be high and incremental
> small but if the system is doing small modifications to blocks scattered
> through the database then I'd expect the opposite - larger incremental vs
> small redo.
>
> Still ...
> I'm interested in real world numbers if anyone has them to share.
>
> Thanks
> Kyle Hailey
> http://db-optimizer.blogspot.com/
>
Other related posts:
