Oracle-L: RE: Lgwr , CKPT - Some Misc. Puzzlers ?

From: DENNIS WILLIAMS <DWILLIAMS_at_LIFETOUCH.COM>
Date: Wed, 19 Feb 2003 07:33:45 -0800
Message-ID: <F001.00551551.20030219073345@fatcity.com>

Vivek
Since I don't see where anyone has responded to your questions, I will make a try and maybe my feeble answers will provoke the more knowledgeable on this list to come to your assistance:

This varies slightly by Oracle version. For Oracle9i, the LGWR writes out the redo log from the redo log buffer when:
A commit record when a user process commits a transaction.
Every 3 seconds.
When the redo log buffer is 1/3 full.
When a DBWn process writes modified buffers to disk, if necessary.
New in 9i, when 1M is reached, which means if the buffer is large, the 1/3 mark will never be reached. I'm not sure if SMON has anything to do with it.
Oracle metadata is stored in the dba_, all_ and user_ tables. Changes to metadata are logged so that if you add a column to a table, for example, you can recover the database and at the appropriate time the metadata change will be re-performed. As was discussed a little while ago on this list, these changes are not performed in a transaction.
http://www.ixora.com.au/q+a/files.htm mentions the reason for the regular writes to the control file is "to update the checkpoint progress record for the thread in the controlfile". My guess is that it has something to do with maintaining recoverability. If your database gets zapped and the control file hadn't been updated in the last 30 minutes, it might be difficult to recover to the last transaction. Again, I am only guessing here. My guess is that 3 seconds is a compromise between no regular interval and "continually".
RAID levels are numbered according to the configuration style, with no relationship to the number of disks involved, except indirectly. You can only create some configurations with a minimum number of disks. For example, you couldn't create a valid RAID 5 configuration with just 2 disks. I believe most of the RAID configurations can be comprised of more than the minimum number of disks. http://www.iomega.com/support/documents/11242.html

Dennis Williams
DBA, 40%OCP, 100% DBA
Lifetouch, Inc.
dwilliams_at_lifetouch.com

-----Original Message-----
Sent: Monday, February 17, 2003 9:19 AM
To: Multiple recipients of list ORACLE-L

Any Brief info on following :-

Qs lgwr writes the redo log buffer to disk when posted by the smon process under what conditions ?

Qs What is metadata ? How are Changes to metadata managed in the log_buffer ?

Qs Why does the Checkpoint process (CKPT) maintains a heartbeat with the control file at a periodic frequency of every 3 seconds ?

Qs The number of Disks "n" in a stripe volume should preferably be a power of 2 namely (2,4,8,16 and so on) for Oracle Databases . For RAID levels 3, 5 and 7, the number of disks should preferably be n+1 . Why ?

--

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

Author: VIVEK_SHARMA
INET: VIVEK_SHARMA_at_infosys.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).
--

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

Author: DENNIS WILLIAMS
INET: DWILLIAMS_at_LIFETOUCH.COM

Fat City Network Services    -- 858-538-5051 http://www.fatcity.com
San Diego, California        -- Mailing list and web hosting services
---------------------------------------------------------------------