From: | "Scott Marlowe" <scott(dot)marlowe(at)gmail(dot)com> |
---|---|
To: | "Thomas Kellerer" <spam_eater(at)gmx(dot)net> |
Cc: | pgsql-general(at)postgresql(dot)org |
Subject: | Re: MVCC and index-only read |
Date: | 2008-11-18 19:57:17 |
Message-ID: | dcc563d10811181157q21ca09a3yadbdc2cd362adfc8@mail.gmail.com |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-general |
On Tue, Nov 18, 2008 at 12:33 PM, Thomas Kellerer <spam_eater(at)gmx(dot)net> wrote:
> If all the columns from the select list are available in the index, then
> Oracle will always prefer the index scan over a table scan (at least I have
> never seen something else). Even for a SELECT that returns all rows of the
> table.
>
> They are taking this concept even further with index organized tables, where
> no real "table data" exists, everything is stored in the index (quited nice
> for e.g. link tables that only consist of two or three integer columns)
Sounds like they're borrowing the code from innodb that does much the
same thing. In Innodb, if a field is indexed, it lives only as an
index, not in the table and an index at the same time.
From | Date | Subject | |
---|---|---|---|
Next Message | Jonah H. Harris | 2008-11-18 19:58:57 | Re: MVCC and index-only read |
Previous Message | Thomas Kellerer | 2008-11-18 19:33:28 | Re: MVCC and index-only read |