Re: Fragmentation X Buffer Cache

Hi.

 Another possibility that you have a lot of FULL TABLE  scans and after exp/imp the rows are sorted better.

 What is your hit ratio anyway ( before and after  exp/imp )?

 Good luck. Michael.

In article <37947B5F.38D3_at_yahoo.com>,
  connor_mcdonald_at_yahoo.com wrote:
> Alexandre Wagner Agostini wrote:
> >
> > Hi folks!
> >
> > I'd like to know what's the relation (if any) between the
fragmentation
> > of the database and buffer cache hit ratio. I'm asking 'cause in the
> > database I'm administering the hit ratio only got raised after the
> > defragmentation. None of the other measures taken, like indexing and
> > buffer cache resizing seemed to work.
> >
> > Thanks already!
> > Wagner
>
> If you used exp/imp then the rows will be reloaded back into the
> database "packed" ie free space within blocks will be all nicely used
up
> (barring pctfree of course)...
>
> Thus you get "more value" for your blocks - that is, each one will
> probably contain more rows, thus increasing the likelihood of a cache
> hit, thus increasing your hit ratio...
> --
> ===========================================
> Connor McDonald
> "These views mine, no-one elses etc etc"
> connor_mcdonald_at_yahoo.com
>
> "Some days you're the pigeon, and some days you're the statue."
>

Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't. Received on Fri Jul 23 1999 - 16:36:55 CDT