Re: Combo xids

From: Heikki Linnakangas <hlinnakangas(at)vmware(dot)com>
To: Simon Riggs <simon(at)2ndQuadrant(dot)com>
Cc: PostgreSQL-development <pgsql-hackers(at)postgresql(dot)org>
Subject: Re: Combo xids
Date: 2013-06-01 18:25:16
Message-ID: 51AA3C8C.6090600@vmware.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

On 01.06.2013 11:22, Simon Riggs wrote:
> What is the benefit?
> Merging xmin/xmax would save 4 bytes per row. On servers with 8 byte
> word length, that means that we'd save 8 bytes per row for tables that
> have between 9 and 40 columns. Which is the majority of tables.

Hmm, it would probably be much easier to squeeze, say, one byte from the
tuple header, than full four bytes. Then you could store store a null
bitmap for upto 16 columns, without crossing the 24 byte mark. That
would already get you much of the benefit.

- Heikki

In response to

  • Combo xids at 2013-06-01 08:22:05 from Simon Riggs

Responses

Browse pgsql-hackers by date

  From Date Subject
Next Message Simon Riggs 2013-06-01 18:34:15 Re: Combo xids
Previous Message Simon Riggs 2013-06-01 18:11:06 Re: detecting binary backup in progress