RE: normalization

From: David Wendelken <davewendelken_at_earthlink.net>
Date: Tue, 1 Nov 2005 21:45:21 -0500
Message-ID: <001601c5df57$7866d0c0$6601a8c0@davidwendelken>

>
>
>BEFORE ROW triggers also need to lock each target row, even if
>nothing is done. This doubles the number of current mode
>buffer gets and increases redo generation significantly.
>

Given that the reason the before row trigger fires is because we are FOR SURE about to change the row, I would expect that we have to pay for the lock ANYWAY...

--
http://www.freelists.org/webpage/oracle-l

Received on Tue Nov 01 2005 - 20:40:49 CST