Re: any over head with DBMS_STATS.ALTER_DATABASE_TABLE_MONITORING

From: Barbara Baker <barbarabbaker_at_yahoo.com>
Date: Thu, 22 Jan 2004 09:39:26 -0800
Message-ID: <F001.005DDE73.20040122093926@fatcity.com>

Too bad Steve Adams' site is not available, cuz that's the place to be.

He says it better than I can, so I've appended a bit of info from him. I'd suggest getting to Steve's site as soon as it's up.

We have monitoring enabled on our 9.2.0.4 database, but it's not currently heavily used. But so far, it's been great.

>From Steve Adam's July 2000 newsletter

(http://www.ixora.com.au/newsletter/2000_07.htm, line spacing below is mine):
Despite the potential for improved statistics gathering, many DBAs have not yet adopted modification monitoring. One of the concerns that people have is that the monitoring might have a significant performance overhead. In fact, this is not the case. The modification counts are maintained in an efficient hash table is the SGA, and are updated without the protection of a latch (although the structure of the hash table itself is protected by the hash table modification latch). Even in heavy OLTP environments, the cost of maintaining the modification counts is likely to be less than 1% of additional CPU usage. However, because of the latch-free nature of the feature, the modification counts are not guaranteed to be accurate. Another source of potential inaccuracy is that if a transaction is rolled back, its changes to the modification counts are not rolled back as well. These inaccuracies have been allowed by Oracle to keep the performance overhead of this feature minimal. Therefore, you can use modification monitoring with confidence that it will not affect performance significantly, while giving you a very helpful indication of which tables may have stale statistics. "

• ryan.gaffuri_at_cox.net wrote:
  > Especially on high transaction tables? I dont have
  > any numbers for transactions/second since we are not
  > live. Any known issues? how does the monitoring
  > work? Does it use an internal trigger and then use
  > SQL to write the data? or does the monitoring data
  > bypass the SQL layer and get written directly to the
  > block?
  >
  > --
  > Please see the official ORACLE-L FAQ:
  > http://www.orafaq.net
  > --
  > Author: <ryan.gaffuri_at_cox.net
  > INET: ryan.gaffuri_at_cox.net
  >
  > Fat City Network Services -- 858-538-5051
  > http://www.fatcity.com
  > San Diego, California -- Mailing list and web
  > hosting services
  >
  

  ---

  
  > 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).
  

  ---

Do you Yahoo!?
Yahoo! SiteBuilder - Free web site building tool. Try it! http://webhosting.yahoo.com/ps/sb/

-- 
Please see the official ORACLE-L FAQ: http://www.orafaq.net
-- 
Author: Barbara Baker
  INET: barbarabbaker_at_yahoo.com

Fat City Network Services    -- 858-538-5051 http://www.fatcity.com
San Diego, California        -- Mailing list and web hosting services
---------------------------------------------------------------------
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 Thu Jan 22 2004 - 11:39:26 CST