Re: v$sql 'db block changes'

  • From: "Jonathan Lewis" <jonathan@xxxxxxxxxxxxxxxxxx>
  • To: <oracle-l@xxxxxxxxxxxxx>
  • Date: Fri, 16 Dec 2005 10:54:09 -0000


If you are thinking about which statements result in the most redo and undo being generated, then it might help to look at statements which are inserts, updates, and deletes, checking the number of rows processed.

This isn't a perfect check, of course, because it
doesn't differentiate the statement that modifies
10 indexes for an insert from the statement that
modifies one index on the insert. But at least is
narrows down the set you have to examine by
eye.

select
   rows_processed, buffer_gets, sql_text
from
   v$sql
where
   command in (
       2,  /* inserts */
       6, /* updates */
       7 /* deletes */
   )
and    rows_processed > {something}
;

Regards

Jonathan Lewis

http://www.jlcomp.demon.co.uk/faq/ind_faq.html
The Co-operative Oracle Users' FAQ

http://www.jlcomp.demon.co.uk/cbo_book/ind_book.html
Cost Based Oracle: Fundamentals

http://www.jlcomp.demon.co.uk/appearances.html
Public Appearances - schedule updated 29th Nov 2005

----- Original Message ----- From: "Allen, Brandon" <Brandon.Allen@xxxxxxxxxxx>
To: <oracle-l@xxxxxxxxxxxxx>
Sent: Thursday, December 15, 2005 5:51 PM
Subject: v$sql 'db block changes'



Anybody have a good method for tracking top SQL by 'db block changes'? It seems this statistic is only tracked by session and instance, but not by SQL statement. The 'disk_reads' are tracked in v$sql and although it wouldn't make sense to track 'disk_writes' (since physical writes are performed in the background as needed by DBWR, not as a direct result of any particular SQL statement), it seems like it would make sense to track 'db block changes' by SQL, doesn't it? I've got heavy writes occurring on a RAID 5 system and it's killing performance, but I'm having trouble identifying the offending SQL. Any ideas?


Thanks,
Brandon

Privileged/Confidential Information may be contained in this message or attachments hereto. Please advise immediately if you or your employer do not consent to Internet email for messages of this kind. Opinions, conclusions and other information in this message that do not relate to the official business of this company shall be understood as neither given nor endorsed by it.



--
//www.freelists.org/webpage/oracle-l


Other related posts: