How large should a redo log file be?

From: Li Chen <lichen_at_cs.buffalo.edu>
Date: 1996/07/03
Message-ID: <4reuc5$92m_at_azure.acsu.buffalo.edu>#1/1


Hi, everyone:

I have an old system running Oracle 7.1.4 on HPUX 9.x and we want to upgrade to Oracle 7.1.6 on HPUX 10.01. The old database is running on noarchivelog mode now and the size of one redo log file is 20M. In addition, every day it is generating 15-20 log files if I check the v$log_history views. We decide to run the new database in archive log mode but I have to decide how large a redo log file should be. 20M seems to too large for redo log file and log_checkpoint_interval is set to 40960 right now. The database is about 15G running Oracle financials on top of it.

Does anyone have an opinion how large a redo log file should be and what value the parameter log_checkpoing_interval should be set?

I appreciate your help.

David

-- 
David Li Chen
Oracle DBA/UNIX System Administrator
Miller Freeman Inc.
dchen_at_mfi.com
Received on Wed Jul 03 1996 - 00:00:00 CEST

Original text of this message