On Mon, 6 Oct 2003 08:54:27 +0100 (BST)
Philip Hazel <ph10@???> wrote:
> On Sun, 5 Oct 2003, Avleen Vig wrote:
>
> > You really neem to do tests, where you need to *write* a lot to the
> > database, to see how the locking effects performance. Reading for the
> > database can happen concurrently but for fallbacks, or other servers
> > which queue mail, locking to write will cause a delay.
>
> Indeed.
But only if it's writing to the same record, plus the time to create the
lock, of course.
Jerry
--
Jerry Jorgenson
jerry@???
http://www.j3iss.com/
