On Fri, Jan 05, 2001 at 04:22:24AM -0800, Michael Cope wrote:
> We're interested in replacing our fallback MX host (sendmail based) with
> an exim based solution and were wondering if anyone had experience with a
> large volume fallback MX using exim.
>
> To give you some idea of the numbers, our current fallback MX processes
> around 400k unique messages a day. Currently (off peak) we have several
> queues with a total of 80k messages (1.5 GB). Our current hardware is a
> Sun E3000 with 4 CPUs and 1.5GB of RAM with queues on a Netapp 630 filer.
Hey, good old NetApp... that's what we use as well.
I am surprised you have 400k messages that cann't be delivered
intantaously on the net. Even we, with 3 million users don't get that
much. What's your setup?
> Of main interest to us is whether the split_spool_directory will be enough
> to balance the queues (ideally fewer than 4000 files per directory) ,
> and what options there are for adding additional flushing queues.
Yes it will.
> We also would be interested in any tools people have to identify and
> remove spam. Remember, the less spam we send the less you get :-).
RBL and DUL lists are good. Otherwise, you are back to clever filter
files and humans.
> Any other experiences people would like to share about Exim in a
> large volume environment is welcome.
Sure, that's what we do...
What exactly do you want to know?
--
Dr Yann Golanski Senior Developer
Please use PGP: http://www.kierun.org/pgp/key-planet
