RE: DBWR Mystery

From: Steve Adams <steve.adams_at_ixora.com.au>
Date: Thu, 31 Aug 2000 19:37:22 +1000
Message-Id: <10605.115943@fatcity.com>

Hi Gopal,

Most of the questions you have asked here were covered in my OpenWorld presentation from 1997 called "The Mysteries of DBWR Tuning". It was based on 7.3 and is way out of date, but it is better than nothing. With that disclaimer in mind, you can get it from http://www.ixora.com.au/tips/mystery.doc.

@ Regards,
@ Steve Adams
@ http://www.ixora.com.au/
@ http://www.christianity.net.au/

@ Going to OpenWorld?
@ Catch the Ixora performance tuning seminar too!
@ See http://www.ixora.com.au/seminars/ for details.

-----Original Message-----
From: K Gopalakrishnan [mailto:kaygopal_at_yahoo.com] Sent: Thursday, 31 August 2000 17:46
To: oracle-l_at_fatcity.com
Cc: oracledba_at_lazydba.com
Subject: DBWR Mystery

I am getting confused with the following DBWR related TAGs .

50 kcbswc DB writer IO clump
16 kcbscc DB writer checkpoint clump
13 kcbsdl DBWR blocks to scan lowest value 50 kcbsdh DBWR blocks to scan highest value 5 kcbsdi DBWR scan depth increment
50 kcbsfs foreground blocks to scan looking for free 200 kcbnbh number buffer headers
247 kcbnbf number buffer objects

Total number of buffers =200 and the DBWR Write batch size is 50. (i.e IO Clump).. But there are two more otehr tags (kcbsfs and kcbsdh) are also 50. Does that mean DBWR will scan so buffers (kcbnbh/4) all the times whenever you need a free buffer in the hash chain. And what is the difference between kcbscc and kcbsfs.(both are kcbnbh/4 here).

And the other point of confusion here is the difference in checkpoint clump and IO clump. What parameter (or algorithm ) used in deciding the kcbsdl and kcbscc .

I am using Oracle 8.0.3 with no buffer pools Received on Thu Aug 31 2000 - 04:37:22 CDT