Re: Index clustering factor

yong321_at_yahoo.com (Yong Huang) wrote in message news:<b3cb12d6.0206202018.58660873_at_posting.google.com>...
> dyou98_at_aol.com (D.Y.) wrote in message news:<f369a0eb.0206200843.5680d099_at_posting.google.com>...
> > versa. As far as
> > performance, I normally don't see a big improvement by tuning
> > clustering factor, not
> > more than a few percent. It could be because we have so much cache
> > these days. You could
>
> Have you done any test on a table that is scanned by parallel slaves,
> in which case Oracle buffer cache is allegedly bypassed? If that test
> still doesn't show difference, make sure you enabled filesystem direct
> I/O or used raw disks and test again.
>

This seems to be another senario where one can benefit from a good clustering factor. It's worth a try when the right situation arises.

> Yong Huang
Received on Fri Jun 21 2002 - 16:29:00 CDT