Oracle FAQ Your Portal to the Oracle Knowledge Grid
HOME | ASK QUESTION | ADD INFO | SEARCH | E-MAIL US
 

Home -> Community -> Usenet -> c.d.o.server -> REPOST: Re: MONITORING

REPOST: Re: MONITORING

From: Brian Tkatch <SPAMBLOCK.Maxwell_Smart_at_ThePentagon.com.SPAMBLOCK>
Date: Sun, 27 Jan 2002 01:10:57 GMT
Message-ID: <4$--$%%%%----_%$$$@news.noc.cabal.int>


On Fri, 25 Jan 2002 21:25:44 +0000, Connor McDonald <connor_mcdonald_at_yahoo.com> wrote:

>> Just to answer my own question. It appears that the 10% rule is
>> followed, using INSERTs and DELETEs, but ignoring UPDATEs. I have a
>> table with, using the query above, has a Percent Added/Deleted of
>> 24%, but Percent Changed of 32.35%. It was not reanalyzed by the
>> GATHER STALE. Anything above 10% Added/Deleted was however.
>>
>> Brian

The 24 should have been .24.

>
>This would make sense since the main emphasis for re-analyze is for when
>row counts have changed. Of course, spurious updates could dramatically
>change the data distribution but I would say that this would not
>typically be the norm.

I figured the same, except for INDEXed columns, in which case UPDATEs are a much larger issue.

>
>Thanks for posting your findings...something people do not do enough of

Happy to. I'd post more, but I'm afraid of showing more of what I don't know, rather than what I figure out. I may think I found something enlightening, only to have my ignorance in the area pointed out.

Brian

This message was cancelled from within The Unacanceller's glorious new software, Lotus 1-2-3 For Rogue Cancellers. Received on Sat Jan 26 2002 - 19:10:57 CST

Original text of this message

HOME | ASK QUESTION | ADD INFO | SEARCH | E-MAIL US