From: | Carol Walter <walterc(at)indiana(dot)edu> |
---|---|
To: | Jim Nasby <decibel(at)decibel(dot)org> |
Cc: | "Spiegelberg, Greg" <gspiegelberg(at)isodxsolutions(dot)com>, <pgsql-admin(at)postgresql(dot)org> |
Subject: | Re: duplicate key violates unique constraint |
Date: | 2007-05-11 13:23:38 |
Message-ID: | E3E7F72F-A86C-437E-A41D-EFD295195FD8@indiana.edu |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-admin |
Yes, I've seen "serial" used in other places than the primary key,
but I was referring to Greg's output. The data type for the sequence
is listed as "bigint". My thought was that the data type was listed
as "serial" for sequences. I just looked at my database and I was
wrong about that. It does list the data type as an integer when you
"describe" the table.
Carol
On May 10, 2007, at 5:17 PM, Jim Nasby wrote:
> On May 10, 2007, at 9:01 AM, Carol Walter wrote:
>> And this is more than a question than an answer, but I thought
>> that sequences had to designated as data type "serial."
>
> Nope... sequences are actually completely un-related to tables. You
> can define a sequence that's not referenced by any table. The
> serial data type is essentially a 'macro' that:
>
> Creates a sequence
> Sets the default value for the field to be the nextval() of that
> sequence
> Sets the field to be NOT NULL
> --
> Jim Nasby jim(at)nasby(dot)net
> EnterpriseDB http://enterprisedb.com 512.569.9461 (cell)
>
>
From | Date | Subject | |
---|---|---|---|
Next Message | Spiegelberg, Greg | 2007-05-11 13:33:22 | Re: duplicate key violates unique constraint |
Previous Message | Steve Holdoway | 2007-05-11 01:10:27 | Re: upgrade 8.0.3 -> 8.2.4 |