Re: Oracle Redo log Switch and clear wait

From: Bastorff <nospam_at_localhost.zz>
Date: Sun, 24 Sep 2000 00:57:48 +0200
Message-ID: <fAaz5.13107$Fl2.113036@nntpserver.swip.net>

"Sudip Sarkar" <sudip_sarkar_at_123india.com> wrote in message news:srv0875vct631_at_corp.supernews.com...
> Hi! All
> While submitting a quite a big transaction process my database justs hangs
> up for about 4to 6 minutes. After monitoring the databse using Precise SQL
> tool I found out that the database is waiting for "Redo log Switch and
> clear wait" for quite a long time. I have tried various combination values
> of redo_log_buffer and log_checkpoint_interval parameters but the problem
> is still nagging. The redo log file size is 30 M and the total size of the
> database is about 10GB.

Did you try a much lower buffer size? 30M is very large, in my opinion. Have you looked into LGWR I/O performance? Are redo log files, data files, rollback segments on the same disk?

Regards,
/fad

> I am really in a fix as I have to finish the benchmarking ASAP.
> Thanx in advance.
>
> --
> Posted via CNET Help.com
> http://www.help.com/
Received on Sat Sep 23 2000 - 17:57:48 CDT