Re: Searching MOS Note on dynamic sampling

From: Jonathan Lewis <jonathan_at_jlcomp.demon.co.uk>
Date: Fri, 28 Feb 2020 10:23:15 +0000
Message-ID: <LNXP265MB15623FB611E70D87C497932FA5E80_at_LNXP265MB1562.GBRP265.PROD.OUTLOOK.COM>


Lothar,

You might be thinking of the hypothesis I raised in an older note to Oracle-L: https://www.freelists.org/post/oracle-l/Re-Re-Re-segment-covers-more-blocks-than-needed,1

Regards
Jonathan Lewis



From: oracle-l-bounce_at_freelists.org <oracle-l-bounce_at_freelists.org> on behalf of Lothar Flatz <l.flatz_at_bluewin.ch> Sent: 28 February 2020 09:46
To: Ghassan Salem
Cc: oracle-l_at_freelists.org
Subject: Re: Searching MOS Note on dynamic sampling

Hi ,

thanks Ghassan. Could not find a blog either. What puzzels me is that we have good search criteria and still don't find anything. That level of detail looks like a MOS note. The details match with our experience and seem to make sense. Could it be a note that was unpublished?

Regards

Lothar

Am 28.02.2020 um 10:21 schrieb Ghassan Salem: Hi Lothar,
I could not find it either. There are several bugs related to how dynamic sampling (for the sample clause) is done, but none mentions this thing.

It is probably in some blog.

Regards

On Fri, Feb 28, 2020 at 9:23 AM Lothar Flatz <l.flatz_at_bluewin.ch<mailto:l.flatz_at_bluewin.ch>> wrote: Hi Guys,

I colleague of mine told me that he read a MOS Note that was describing the details of the way dynamic sampling is done. The note was mentioning that only bitmap blocks are sampled and blocks less than 25% full are considered empty. In that same note also deferred segment creation is mentioned. As my colleague is very experienced I believe is memory is correct. However, we cannot find that note any more.

Can somebody point us to the doc?

Regards and thank

Lothar

--

--

--

http://www.freelists.org/webpage/oracle-l Received on Fri Feb 28 2020 - 11:23:15 CET

Original text of this message