Re: Slow performance on index creation

From: Thomas J Kyte <tkyte_at_us.oracle.com>
Date: 1996/03/22
Message-ID: <4it36t$fu8_at_inet-nntp-gw-1.us.oracle.com>#1/1

"Brian M. Biggs" <bbiggs_at_cincom.com> wrote:

<snip>

>Our redo logs are only 512K. I know this is small, but I didn't think
>it would have this much of an effect on performance. Could this be some
>of the reason for our lousy numbers? Is Oracle spending a lot of time
>doing log switches and such?

Increase your logs immediately. You probably have the 2 default 512k log files. The index creation will generate logs and will whip thru them very fast. Look at your $ORACLE_HOME/rdbms/log/alert.ora file. You will probably see lots of 'Cannot allocate new log, checkpoint not complete'. These are performance killers. Add lots of log (50 meg would be good during your index creation). You can always add / remove logs after the fact. For the index creations I would recommend a couple of good sized log files. For example, I run with 2 25 megs logs myself. The goal is to remove the 'Cannot allocate new log' messages from your system.

<snip>

>Regards,
>Brian
 

>--
>Brian M. Biggs mailto:bbiggs_at_cincom.com
>Cincom Systems, Inc. voice: (513) 677-7661
>http://www.cincom.com/

Thomas Kyte
tkyte_at_us.oracle.com
Oracle Government

---

opinions and statements are mine and do not necessarily reflect the opinions of Oracle Corporation. Received on Fri Mar 22 1996 - 00:00:00 CET