Re: update performance

Krishna,   

   What do you mean by 'an index on the two columns' ? There is an index on each column ? Or both columns belong to a concatenated index ? If the latter case is the right one, it should be on (colA, colC), not the reverse. I think that the first thing to do is EXPLAIN PLAN, anyway. Beware of triggers, too.

HTH, Stephane Faroult
Oriole Corporation

>RE: Date comparison questionI ahve a requirement to update a table using a
>joing - details:
>Table A: (colA, colB,...): 2 million rows
>Table B: (colA, colC,...): 2 milion rows
>
>update cmd:
>
>update tableA A
>set colA = (select B.colC fdrom tableB B where B.tableB.colA =
>A.tableA.colA )
>
>I tried it in PL/SQL and as the straight SQL but it takes FOREVER without
>coming back...
>should I be doiing something better?
>
>(I have an index on the two columns on TableB)
>
>Thanks
>-Krishna.
>

--
This mail was sent through http://www.mail-inspector.de
Mail Inspector is a free service of http://www.is-fun.net CHECK OUT !

This mail was sent from host 194.3.142.126