Thanks
to all of you on this issue...
-Original Message-From: Gogala, Mladen
[mailto:[EMAIL PROTECTED]]Sent: Wednesday, April 24, 2002 8:12
PMTo: Multiple recipients of list ORACLE-LSubject: RE:
LOG_BUFFER Parameter Question
Alternatively, you size it until
Yep,
wrong tree. size it like: redo blocks per tx * tx/sec * 3/2 *physical
blocksize>
Anjo.
Denham Eva wrote:
Hello,
I have been wondering about how to
work out what the optimal LOG_BUFFER parameter,
compared to the physical size of the
REDOLOG file, for the best optimization and writes from
log buffer gets written
a) every 3 seconds
b) every commit (more or less)
c) 1/3 full
d) 1M full
thus its unrelated to redo log file size, and in most
cases unlikely that anything above 512k-1m is going to
give benefits. Check for 'log buffer waits' in your
statspack data.
hth
connor
---
Title: LOG_BUFFER Parameter Question
Are you seeing any significant times from
wait-events related to redo log buffering?
- Original Message -
From:
Denham Eva
To: Multiple recipients of list ORACLE-L
Sent: Tuesday, April 23, 2002 11:58
PM
Subject: LOG_BUFFER
Alternatively, you size it until redo_log_space_wait
goes away from v$system_events.
-Original Message-From: Anjo Kolk
[mailto:[EMAIL PROTECTED]]Sent: Wednesday, April 24, 2002 4:59
AMTo: Multiple recipients of list ORACLE-LSubject: Re:
LOG_BUFFER Parameter QuestionYep,
On Wed, Apr 24, 2002 at 10:12:22AM -0800, Gogala, Mladen wrote:
Alternatively, you size it until redo_log_space_wait goes away from
v$system_events.
it is v$system_event
-Original Message-
Sent: Wednesday, April 24, 2002 4:59 AM
To: Multiple recipients of list ORACLE-L