> why I/O is slow).
> > >
> > > So your solution of going with the INDEX_FFS
> hint is
> > > too reduce the number
> > > of
> > > I/O. Well what if the problem is that the I/O is
> too
> > > slow ?
> > >
> > > Anjo Kolk
>
ce the number
> > of
> > I/O. Well what if the problem is that the I/O is too
> > slow ?
> >
> > Anjo Kolk
> >
> > http://www.oraperf.com
> >
> >
> > Robertson Lee - lerobe wrote:
> >
> > > All,
> > >
> > > Oracle 8.0.5.0.0
hint is
> too reduce the number
> of
> I/O. Well what if the problem is that the I/O is too
> slow ?
>
> Anjo Kolk
>
> http://www.oraperf.com
>
>
> Robertson Lee - lerobe wrote:
>
> > All,
> >
> > Oracle 8.0.5.0.0 Tru64 4.0f
> >
>
We are having a performance issue with one of our applications and it
looks
> like it is pointing to the db file sequential reads event. I have been
doing
> some digging around and a recommendation in a book I have has suggested a
> possible way to alleviate this is to use the INDEX_FFS hin
ointing to the db file sequential reads event. I have been doing
some digging around and a recommendation in a book I have has suggested a
possible way to alleviate this is to use the INDEX_FFS hint in the query.
Anyone used this before ??. or indeed any other recommendations where I
should be looking
the I/O is too slow ?
Anjo Kolk
http://www.oraperf.com
Robertson Lee - lerobe wrote:
> All,
>
> Oracle 8.0.5.0.0 Tru64 4.0f
>
> We are having a performance issue with one of our applications and it looks
> like it is pointing to the db file sequential reads event. I have been do
All,
Oracle 8.0.5.0.0 Tru64 4.0f
We are having a performance issue with one of our applications and it looks
like it is pointing to the db file sequential reads event. I have been doing
some digging around and a recommendation in a book I have has suggested a
possible way to alleviate this is
events in
v$system_event for
'buffer busy
waits' (2 million)
and mainly
'db f
$waitstat showed
me lots of waits for 'data blocks', v$system_event showed me waits for 'db
file sequential reads' and 'buffer busy waits'. From v$session_wait, I
tracked this down to a single table. v$filestat shows me lot of physical
reads for the disk where the dataf
on "db file sequential reads" (85% of the active
waits)
No disk I/O problem is detected.
Buffer cache hit_ratio = 92.30 %
Is it normal to have such amount of waits on "db file sequential reads" ?
SVRMGR> select n1.event "Event Name",
2>
10 matches
Mail list logo
