Re: Bulkloading using COPY - ignore duplicates?

On Thu, Dec 13, 2001 at 01:25:11PM +0000, Lee Kindness wrote:
> That's what I'm currently doing as a workaround - a SELECT DISTINCT
> from a temporary table into the real table with the unique index on
> it. However this takes absolute ages - say 5 seconds for the copy
> (which is the ballpark figure I aiming toward and can achieve with
> Ingres) plus another 30ish seconds for the SELECT DISTINCT.

Then your column really isn't unique, so how about dropping the unique index,
import the data, fix the duplicates, recreate the unique index - just as
another possible work around ;)

Patrick