From: | Scott Marlowe <smarlowe(at)g2switchworks(dot)com> |
---|---|
To: | Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> |
Cc: | Ben <bench(at)silentmedia(dot)com>, pgsql general <pgsql-general(at)postgresql(dot)org> |
Subject: | Re: index creation on 7.3 |
Date: | 2007-01-10 17:40:27 |
Message-ID: | 1168450827.20602.187.camel@state.g2switchworks.com |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-general |
On Tue, 2007-01-09 at 23:12, Tom Lane wrote:
> Scott Marlowe <smarlowe(at)g2switchworks(dot)com> writes:
> > On Tue, 2007-01-09 at 17:17, Ben wrote:
> > If I create an index on a table that needs to be vacuumed in 7.3, will the
> > dead rows get indexed as well?
>
> > Yes. I think. I'm pretty sure that's still true in 8.2. as well. The
> > way indexes work is that the index points to the oldest version of a
> > tuple in existence, said tuple which then points to the next newer
> > version and so on.
>
> IIRC, CREATE INDEX will ignore tuples that are "definitely dead", ie,
> too old to be visible to any open transaction. It *will* (and must)
> index tuples that are recently dead but might still be visible to some
> open transaction.
So, it depends on one's definition of dead...
I don't want to go on the cart...
From | Date | Subject | |
---|---|---|---|
Next Message | A. Kretschmer | 2007-01-10 17:41:04 | Re: Foreign Key Identification |
Previous Message | Tom Lane | 2007-01-10 17:17:25 | Re: Foreign Key Identification |