| Oracle FAQ | Your Portal to the Oracle Knowledge Grid | |
 |
 | |||
| Oracle FAQ | Your Portal to the Oracle Knowledge Grid | |
|
| |||
Home -> Community -> Mailing Lists -> Oracle-L -> Re: db_block_checkpoint_batch
Riyaj,
Thanks for your response.
The thing that puzzles me is that the seconds_in_wait time for all slaves dbwr processes equals to the time since the database was started. I would expect the seconds_in_wait time for slaves dbwr to be less than the database uptime. Does that mean that all slaves processes are idle? Or is the seconds_in_wait time for slaves dbwr interpreted differently than the rest of the db sessions?
I get conflicting opinion that Oracle does not support asyncio on Unix filesystem. Any idea if that is true?
Thanks.
--Elain
>From: Riyaj_Shamsudeen_at_i2.com
>Reply-To: ORACLE-L_at_fatcity.com
>To: Multiple recipients of list ORACLE-L <ORACLE-L_at_fatcity.com>
>Subject: Re: db_block_checkpoint_batch
>Date: Tue, 13 Feb 2001 09:16:33 -0800
>
>
>Hi elain
> I have not personally encountered performance degradation setting
>this
>to high value. But I will attempt to help you.
> DBWR process is waiting for the IPC messages from the background
>process or from the server processes, for the requests to write the dirty
>buffers. That's how the processes communicate among themselves and wiating
>for an IPC message is normal. Also DBWR is the process which scans the
>buffer cache to create a list of buffers to write and then it distributes
>that list to the slave process to write. So slave process has to wait for
>the null event during no work. So, this is fine too.
> db_block_checkpoint_batch determines what portion of db writes can be
>used for the slow checkpoints (in 7.3). By setting this to high value, the
>checkpoint will complete faster, but the speed at which DBWR cleans the
>buffer cache may be slightly slower. So more free buffer waits. You may
>have to play around these parameters little bit to determine the optimum
>value for your environment. I would set checkpoint batch size to be around
>1/8 th or 1/10th of Internal batch write size and go upwards. Also make
>sure that your log buffer is small enough and the DBW doesn't wait too much
>for the log file sync event. Converting the file system based database to
>raw database and turning on async IO are other options.
> In fact, Steve Adams has written a paper about this DBWR tuning.
>Search for db_block_checkpoint in his website www.ixora.com.au. You will
>find his paper.
>
>Thanks
>Riyaj "Re-yas" Shamsudeen
>Certified Oracle DBA
>"This is my opinion and does not legally bind me or my employer. Use at
>your own risk"
>
>
>
> "elain he"
> <elainhe_at_hotm To: Multiple recipients of
>list ORACLE-L <ORACLE-L_at_fatcity.com>
> ail.com> cc:
> Sent by: Subject:
>db_block_checkpoint_batch
> root_at_fatcity.
> com
>
>
> 02/13/01
> 08:26 AM
> Please
> respond to
> ORACLE-L
>
>
>
>
>
>
>Hi,
>Has anyone seen database performance degradation by setting this to a high
>value?
>
>We set this value to 128 from the default value 4 and noticed a lot more
>free buffer waits event.
>
>All database writers (DB01-10) are waiting for Null Event. The
>seconds_in_wait time from v$session_wait is 129614(~1.5days), ie the same
>time since the database started. The parent DBWR process is waiting on
>rdbms
>ipc message and the seconds_in_wait time is 0.
>
>Here's the configuration:
>DB 7.3.4
>db_writer=10
>async_io disable
>Solaris 2.6/Sun T3 Array
>_db_block_write_batch=512
>db_block_checkpoint_batch=128 (512/4)
>
>Appreciate any help you can provide.
>Thanks.
>
>--Elain
>_________________________________________________________________
>Get your FREE download of MSN Explorer at http://explorer.msn.com
>
>--
>Please see the official ORACLE-L FAQ: http://www.orafaq.com
>--
>Author: elain he
> INET: elainhe_at_hotmail.com
>
>Fat City Network Services -- (858) 538-5051 FAX: (858) 538-5051
>San Diego, California -- Public Internet access / Mailing Lists
>--------------------------------------------------------------------
>To REMOVE yourself from this mailing list, send an E-Mail message
>to: ListGuru_at_fatcity.com (note EXACT spelling of 'ListGuru') and in
>the message BODY, include a line containing: UNSUB ORACLE-L
>(or the name of mailing list you want to be removed from). You may
>also send the HELP command for other information (like subscribing).
>
>
>
>
>--
>Please see the official ORACLE-L FAQ: http://www.orafaq.com
>--
>Author:
> INET: Riyaj_Shamsudeen_at_i2.com
>
>Fat City Network Services -- (858) 538-5051 FAX: (858) 538-5051
>San Diego, California -- Public Internet access / Mailing Lists
>--------------------------------------------------------------------
>To REMOVE yourself from this mailing list, send an E-Mail message
>to: ListGuru_at_fatcity.com (note EXACT spelling of 'ListGuru') and in
>the message BODY, include a line containing: UNSUB ORACLE-L
>(or the name of mailing list you want to be removed from). You may
>also send the HELP command for other information (like subscribing).
-- Please see the official ORACLE-L FAQ: http://www.orafaq.com -- Author: elain he INET: elainhe_at_hotmail.com Fat City Network Services -- (858) 538-5051 FAX: (858) 538-5051 San Diego, California -- Public Internet access / Mailing Lists -------------------------------------------------------------------- To REMOVE yourself from this mailing list, send an E-Mail message to: ListGuru_at_fatcity.com (note EXACT spelling of 'ListGuru') and in the message BODY, include a line containing: UNSUB ORACLE-L (or the name of mailing list you want to be removed from). You may also send the HELP command for other information (like subscribing).Received on Tue Feb 13 2001 - 12:38:50 CST
 |
 |