Re: Transaction Gradually Slows to a crawl! (please help)

Richard Foote wrote:
>
> "Tanel Poder" <tanel@@peldik.com> wrote in message news:<3ed62044$1_1_at_news.estpak.ee>...
> > > Disabling redo logging effectively disables incremental checkpointing.
> >
> > In my experience, checkpointing still occurs during log switches, thus you
> > should create huge redo logs (temporarily) and set log_checkpoint_interval &
> > timeout accordingly (version 8.1.7).
> >
>
> Hi Tanel,
>
> Can you explain to me how you get a log switch when you disable redo
> logging and how huge redo logs help when they're not being written to
> ???
>
> Cheers ;)
>
> Richard

If I remember correctly (translation: could not be arsed to do a check), then all the normal processing takes place, just the physical write by lgwr is not done. Thus you get log 'switches' when a log becomes 'full' and thus get a checkpoint etc etc.

Hence monster-sized log files will assist

hth
connor

-- 
=========================
Connor McDonald
http://www.oracledba.co.uk

"Some days you're the pigeon, some days you're the statue"