Re: AW: Re: AW: Re: amm vs. asmm

oh ok, missread! but I guess the end effect is similar, heavy I/O

Thanks for the clarification

On Fri, Jan 24, 2020 at 5:45 PM Stefan Koehler <contact_at_soocs.de> wrote:

> Hello Cheng,
> FYI - the behavior that is described in the Tweet is kinda different. I
> just copy & paste my reply: "Buffer cache was shrinking from 130 GB to 118
> GB over a time period of one month but at the time when it was 118 GB the
> DB only got 2.077.979 buffer headers (2.077.979 x 8k = 16623832 k = 16234
> M) within the 118 GB."
>
> ... so the buffer cache size was still 118 GB but only round about 16 GB
> of it was used.
>
> Best Regards
> Stefan Koehler
>
> Independent Oracle performance consultant and researcher
> Website: http://www.soocs.de
> Twitter: _at_OracleSK
>
> > Ls Cheng <exriscer_at_gmail.com> hat am 24. Januar 2020 um 11:27
> geschrieben:
> >
> > Hi
> >
> > I have also observed behaviour stated in [
> https://twitter.com/jolliffe/status/1094195271448186881?s=20](https://twitter.com/jolliffe/status/1094195271448186881?s=20
> .)
> >
> > I have seen it when no minimum memory parameters are set. For example I
> have a customer who has a database with 40GB sga_target set, no other
> memory parameters and after 3 or 4 weeks the buffer cache was left with
> 64MB only.
> >
> > So I think the best practice would be setting minimum and sga_target.
> >
> > BR
>

--
http://www.freelists.org/webpage/oracle-l