Re: Unique rows without a key - Mailing list pgsql-general

From GH
Subject Re: Unique rows without a key
Date
Msg-id 20010612181956.B11208@over-yonder.net
Whole thread Raw
In response to Unique rows without a key  (Steve Micallef <stevenm@ot.com.au>)
List pgsql-general
On Wed, Jun 13, 2001 at 09:00:33AM +1000, some SMTP stream spewed forth:
> Hi,
>
> Is there any way to force every row in a table to be unique without
> creating a unique key?

Not exactly what you want, I suppose, but you could create a unique index
on the oid column --- the only additional overhead would be that of the
index.

> My tables are fairly large (growing by a gig a day) and creating a primary
> key would basically double my disk usage.

Obviously, the system would have to have some way of knowing what other
values exist -- you would have to have a unique index on some column.

> Thanks for your help,
> Steve Micallef

gh


pgsql-general by date:

Previous
From: Tom Lane
Date:
Subject: Re: Connection Timeout ?
Next
From: Tom Lane
Date:
Subject: Re: Merge join exhausting swap space