Re: Reduce heap tuple header size - Mailing list pgsql-patches

From Bruce Momjian
Subject Re: Reduce heap tuple header size
Date
Msg-id 200206220357.g5M3vGJ13783@candle.pha.pa.us
Whole thread Raw
In response to Re: Reduce heap tuple header size  (Alvaro Herrera <alvherre@atentus.com>)
Responses Re: Reduce heap tuple header size  (Tom Lane <tgl@sss.pgh.pa.us>)
List pgsql-patches
Alvaro Herrera wrote:
> Tom Lane dijo:
>
> > Jan Wieck <JanWieck@Yahoo.com> writes:
> > > Here are some numbers:
> >
> > > Current CVS tip:     tps  34.1, 38.7, 36.6
> > >                  avg(tps) 36.4
> >
> > > With patch:          tps  37.0, 41.1, 41.1
> > >                  avg(tps) 39.7
> >
> > > So it saves less than 3% disk space at the cost of about 9% performance
> > > loss.
> >
> > Uh ... isn't more TPS better?

9%, that is a dramatic difference.  Is it caused by the reduced disk
space (Jan's numbers are correct) or by the extra overhead in the merged
fields (Jan's numbers are backwards)?  Jan will tell us soon.

> Also, is that 3% in disk space savings the actual number, or just copied
> from the "anawhat"?

The 3% is savings from a sample database.  Header size is 11% reduced.

--
  Bruce Momjian                        |  http://candle.pha.pa.us
  pgman@candle.pha.pa.us               |  (610) 853-3000
  +  If your life is a hard drive,     |  830 Blythe Avenue
  +  Christ can be your backup.        |  Drexel Hill, Pennsylvania 19026

pgsql-patches by date:

Previous
From: Joe Conway
Date:
Subject: show() function
Next
From: Bruce Momjian
Date:
Subject: Re: [ODBC] ODBC Patch to prevent setting of KSQO on 7.3+ servers