Oracle FAQ Your Portal to the Oracle Knowledge Grid
HOME | ASK QUESTION | ADD INFO | SEARCH | E-MAIL US
 

Home -> Community -> Mailing Lists -> Oracle-L -> Re: Do we really need multiple dbwr when AIO is used?

Re: Do we really need multiple dbwr when AIO is used?

From: fengchunpei <biti_rainy_at_itpub.net>
Date: Sun, 18 Apr 2004 23:31:52 +0800
Message-Id: <20040418152520.9F81072C7DF@turing.freelists.org>


hi,zhu chao=A3=A1

=09=09I think if the AIO is available ,then need not multiple dbwr=  . one DBWR can "use" more cpus and disk drivers . when os =  do not supports AIO ,one dbwr can only use a cpu and driver=  (even if there has more) per write. It waste the hard =  resource .

 =09=09        yahoo id: feng_chunpei 
                msn: biti_rainy_at_hotmail.com
=A1=A1=A1=A1=A1=A1=A1=A1=A1=A1=A1=A1=A1=A1=A1=A1 biti_rainy_at_itpub.net
                =A1=A1

>hi, friends:
> From time to time, I see various of notes say, in heavy=
 load, you'd better use multiple dbwr even if AIO is used. I=  think AIO allows dbwr to issue new write request to the driver=  without waiting for the previous write request to complete.
> When disk itself is 100 busy, I think more dbwr won't=
 help.If disk is not 100 busy, it is the oracle dbwr itself won't=  catch up, then we should use multiple dbwr. I did a performance=  test today, comparing the result of using 1 and 4 dbwr with aio=  on.
>Test Environment:
> Sun FireV880. 8*1.2GHZ CPU. 16G memory.
> Storage: HDS 9570.
> Oracle 8.1.7.4. Veritas Volume Manager 3.5. All=
 datafile/redo/control/temp on raw device.
> Redo log size is 500M * 5 group.
> Shared_pool_size is 80M and data buffer is 1.2G.
>
>Build an empty database on raw device, with 25G of freespace.=
 One target tablespace, extent management local uniform size 5m.
>Do a 1.2G dmp table import. No special tuning.
>
>First time: 1 dbwr. import. It took 20 minutes and 30 seconds.
>Second time: drop user, recreate user, configure 4 dbwr process,=
 db_block_lru_latches=3D8.
> do import again..Time is 21Minutes and 44=
 seconds.
>Third time: 1 dbwr. Drop user and recreate. The same parameter=
 as test 1. (I did it as test 1, in case SAN cache was=  relatively empty, so maybe performance better).
> This time the import time is 21Minutes and 38=
 seconds.
>
>There is no performance difference using 1 dbwr or 4 dbwr. I did=
 a statspack, and found during peak time, there were 2500+ write=  per second. And according to the test, one dbwr can handle it=  efficiently. From solaris iostat -xn result, I saw 4000+ write=  per second during peak time(redo write plus dbwr write).
>
> I think import is a heavy task for dbwr, few database=
 will write 2500+ per second(Our database do 200 write per=  second). Since in such a test, one dbwr can handle the request=  efficiently, I wonder whether those guys who claim the need for=  multiple dbwr do have real world experience, when aio is=  avaliable. If yes, please give us some data, not just the rule=  of the numb.
>
>Please comment.
>
>Regards
>Zhu Chao.
>


Please see the official ORACLE-L FAQ: http://www.orafaq.com

To unsubscribe send email to: oracle-l-request_at_freelists.org put 'unsubscribe' in the subject line.
--
Archives are at http://www.freelists.org/archives/oracle-l/
FAQ is at http://www.freelists.org/help/fom-serve/cache/1.html
-----------------------------------------------------------------
Received on Sun Apr 18 2004 - 10:32:50 CDT

Original text of this message

HOME | ASK QUESTION | ADD INFO | SEARCH | E-MAIL US