AW: Is stats update during COPY IN really a good idea? - Mailing list pgsql-hackers

From Zeugswetter Andreas SB
Subject AW: Is stats update during COPY IN really a good idea?
Date
Msg-id 11C1E6749A55D411A9670001FA6879633682E5@sdexcsrv1.f000.d0188.sd.spardat.at
Whole thread Raw
Responses Re: AW: Is stats update during COPY IN really a good idea?  (Bruce Momjian <pgman@candle.pha.pa.us>)
List pgsql-hackers
> We have a TODO item
>     * Update reltuples in COPY
> 
> I was just about to go do this when I realized that it may not be such
> a hot idea after all.

Imho it is not a good idea at all. The statistics are a very sensitive area,
that imho should only be calculated on request. I already don't like the
statistics that are implicitly created during create index.

Eighter you have online stats keeping or you don't.
For me this is a definite all or nothing issue. Anything inbetween is 
only good for unpleasant surprises.

I have very strong feelings about this, because of bad experience.
I would be willing to go into detail.

A syntactic extension to copy ("with analyze") on the other hand would 
be a feature.

Andreas


pgsql-hackers by date:

Previous
From: "Christopher Kings-Lynne"
Date:
Subject: RE: DROP CONSTRAINT patch
Next
From: Denis Perchine
Date:
Subject: Re: Re: AW: Plans for solving the VACUUM problem