In article <6ctnvtcc1uk7ocpvt9qd7pmiejvapu4m28_at_4ax.com>, "Sybrand Bakker"
<oradba_at_sybrandb.demon.nl> wrote:
>>
>>
> A 10M log_buffer is way too big and is just asking for trouble. You
> should also check whether there are any 'checkpoint not complete' or
> 'cannot allocate log' messages in your alert<sid>.log.
> Chances are your online redolog files are way too small (the default is
> 200k which is just ridiculous, an adequate minimum is 4 groups consisting
> of 2 files per group, each file with a size of 5M)
>
> Also locating the complete database on one single disk will definitely
> slow down everything, and it will also increase the chance you loose your
> complete database (I notice you are running in noarchivelog which you
> shouldn't do)
>
> Hth
>
>
>
>
> Sybrand Bakker, Senior Oracle DBA
>
> To reply remove -verwijderdit from my e-mail address
Probably the best advice would be to install statspack, execute a few
snapshots and then try to see what can be improved, based on the statspack
report.
--
Mladen Gogala
Received on Wed Nov 21 2001 - 16:09:03 CST
