Oracle advanced tuning course

Hi list
as you know I'm not an expert, but I think after all this 7 years, and
suffering I had in all hte database we installed, I can give a course to
teach a new dba tecniques to get an advanced level.
I think this is sometimes only to point to the fact some feature exists,
I tried to test any thing before giving an opinion, and investigated about.
For exapmle I couldn't get oracle use histograms I'll do further
investigations.

I have to add several points, but I'll do when I get time, bucause it tookme
a lot of time to write it.

I respect other words (something not common or I can't find in documentation
using other words) , for examle from this lis I copy cary's comment about
RAC, I could change cary's words and write it as if I would had said, but I 
think this is unethic and stupid, because the day someone askme about it
will be obvious that weren't my words. The day I have enough rac expertience
to give my opinion I'll replace it.
I don't know too much about copyright any correction will be welcome, I
suppose to put the link is enough.

This paper will be free available on my site

This is almost the final draft,is far miles away from my previous paper,
this is at least all in english.

Any specific critic is welcome.

 www.geocities.com/juancarlosreyesp/OracleAdvancedTuningCourse.pdf

sorry, my daily bandwidth is short.

Juan Carlos Reyes Pacheco
OCP
-------Original Message-------
 
From: oracle-l@xxxxxxxxxxxxx
Date: 07/26/04 19:10:53
To: oracle-l@xxxxxxxxxxxxx
Subject: Re: 9.2 V$ views
 
> (like even analyzing DD in 9.2, etc.)
>
> Can't that lead us to other problems?
 
Yep, it can ;)
But in 9.2 it's supported and in 10g it's default anyway...
 
Btw, the dba_extents performance problem you mentioned earlier probably
comes from the fact that in LMT configuration you can get detailed extent
information only from the segment header itself, this means at least one
consistent get per segment which you're querying. Given that lots of segment
headers aren't in buffer cache, a count(*) on dba_extents may cause
thousands of physical IO's, recursive calls and latching...
 
SQL> set autot trace stat
SQL> select count(*) from dba_extents;
 
1 row selected.
 
 
Statistics
----------------------------------------------------------
41228 recursive calls
1 db block gets
20799 consistent gets
3389 physical reads
0 redo size
380 bytes sent via SQL*Net to client
511 bytes received via SQL*Net from client
2 SQL*Net roundtrips to/from client
0 sorts (memory)
0 sorts (disk)
1 rows processed
 
So, dba_extents is a quite dangerous view to query in LMT environment.
 
Tanel.
 
 
----------------------------------------------------------------
Please see the official ORACLE-L FAQ: http://www.orafaq.com
----------------------------------------------------------------
To unsubscribe send email to: oracle-l-request@xxxxxxxxxxxxx
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
-----------------------------------------------------------------
----------------------------------------------------------------
Please see the official ORACLE-L FAQ: http://www.orafaq.com
----------------------------------------------------------------
To unsubscribe send email to:  oracle-l-request@xxxxxxxxxxxxx
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
-----------------------------------------------------------------

Other related posts: