Re: Performance issue - high Free buff wait

From: Pap <oracle.developer35_at_gmail.com>
Date: Mon, 18 Jan 2021 18:21:58 +0530
Message-ID: <CAEjw_fjicfbX6m+VyRL38qPG_MJYdWtOsjNfu2TcYTMQPXM8oQ_at_mail.gmail.com>





Confirmed from the infra DBA's , huge pages are not in use here.

My understanding was that there would always be some non zero swapping seen. Will check from the host if it's true what's showing in the OEM. I have attached the IO STATS section of the AWR, the DBWR response does show 100MS+ during issue period VS ~35MS during normal time and its not in ASM its file system only. I am yet to see if this response is matching with the IOSTAT/DSTAT from the host.

Regards
Pap

On Mon, Jan 18, 2021 at 7:26 AM Mladen Gogala <gogala.mladen_at_gmail.com> wrote:

> There is also the question of huge pages. Is the SGA allocated as huge
> pages? All that talk about asynchronous IO leads to conclusion that the
> direct IO is not used ant the system may be double buffered.
>
> On 1/17/21 3:45 PM, Jonathan Lewis wrote:
> > Don't be fooled by the fact I typed MB instead of GB - you have 512GB
> > of memory and a huge fraction of it is "free". Why isn't more given to
> > the database.
>
> --
> Mladen Gogala
> Database Consultant
> http://mgogala.byethost5.com
>
> --
> http://www.freelists.org/webpage/oracle-l
>
>
>



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


  • application/vnd.openxmlformats-officedocument.spreadsheetml.sheet attachment: DBWR_Response.xlsx
Received on Mon Jan 18 2021 - 13:51:58 CET

Original text of this message