you get that number (512k) and where can we find more info
about that.
-Original Message-
From: K Gopalakrishnan [mailto:[EMAIL PROTECTED]]
Sent: Wednesday, October 09, 2002 3:24 PM
To: Multiple recipients of list ORACLE-L
Subject: RE: Table Scans
Jay:
The other
:
yahoo.comSubject: RE: Table Scans
Sent
I am curious about something in your query.
SELECT distinct(o.object_name),o.object_type,o.owner
FROM dba_objects o,x$bh x
WHERE x.obj=o.object_id
AND o.object_type='TABLE'
AND standard.bitand(x.flag,524288)0
AND o.owner'SYS';
Where did you learn of the correct pattern for the
AND
Jay:
The other option is to look for 'db file scattered read' waits
and join with dba_extents/segments to get the segments which are
accessed via full table scan. This would be better alternate since
you don't need to scan the entire buffer cache to get the names of
the segments whose blocks are
http://www.think-forward.com/sql/bufferts.htm
-Original Message-
Sent: Wednesday, October 09, 2002 4:00 PM
To: Multiple recipients of list ORACLE-L
I am curious about something in your query.
SELECT distinct(o.object_name),o.object_type,o.owner
FROM dba_objects o,x$bh x
WHERE
ORACLE-L
Subject: RE: Table Scans
Jay:
The other option is to look for 'db file scattered read' waits
and join with dba_extents/segments to get the segments which are
accessed via full table scan. This would be better alternate since
you don't need to scan the entire buffer cache to get
Title: RE: Table Scans
Mladen,
Look up definition of x$bh.flag.
For example, http://www.jlcomp.demon.co.uk/buf_flag.html
Correction to the query, join should be x.obj=o.data_object_id
And as a note, one needs to take care about clustered tables.
Alex.
-Original Message-
From
Aargh! Thanks! Flag 19 is sequential only access and bitand with 2**19
will get only the buffers belonging to objects accessed using full table scan.
Thanks!
On 2002.10.09 23:08 [EMAIL PROTECTED] wrote:
Mladen,
Look up definition of x$bh.flag.
For example,
