Eric Cox <[EMAIL PROTECTED]> writes on 22 August 2000 at 22:47:43 -0700
>
>
> Bryan White wrote:
> >
> > Sorry if this is a little off topic.
> >
> > I am trying to speed up my outbound mail. I am running a set of ezines and
> > we are currently sending around 3 million emails per day. Currently I have
> > a process that envokes qmail-inject for each message. I am thinking of
> > doing a first attempt delivery inside my program and only if that fails,
> > hand it off to qmail-inject.
>
> Just out of curiosity: is there an obvious reason that I'm missing that
> a custom app would send out mail faster than qmail?
Yes; if you can skip the step of writing to the queue, you save
considerable disk IO and a fork or two. Since something vaguely like
90% of deliveries succeed on the first try, it's been frequently
suggested that trying a delivery before queueing would be a winning
strategy. So, it's not that the custom app would deliver faster than
qmail; it's that it might win to try delivering before queueing, and
only queue if the first delivery attempt fails.
--
Photos: http://dd-b.lighthunters.net/ Minicon: http://www.mnstf.org/minicon
Bookworms: http://ouroboros.demesne.com/ SF: http://www.dd-b.net/dd-b
David Dyer-Bennet / Welcome to the future! / [EMAIL PROTECTED]
