[HLUG] Exim 4.50 on Debian 3.1 external greet_pause?

[John Hedges]

Hi Andrew
 
> >> >Maybe ident (rfc1413)? I have 'rfc1413_query_timeout = 0s' in most
> >> >configs as this has solved delays in the past (especially when using
> >> >exim as a smarthost for a pc without an ident server).
> >> 
> >> Ok did a test on my machine and this is correct as the firewall here
> >> logs the incoming port 13 request.  Where abouts in the config do I
> need
> >> to disable it?  I use the split config.
>  
> >It's in /etc/exim4/conf.d/main/02_exim4-config_options :
> 
> Yes, I found it, as well as reading a load of docs, I have managed to
> implement most of what I want (i.e, TLS, Callout on secondary MX etc).
> The only problem I have now is that it takes around 30 minutes to spool
> mail from the secondary to the primary.  It is not a huge issue, but I
> can't work out why it is doing this?  I specified the domain in the
> relay domains prompt during the configure stage.  I want it to spool as
> soon as possible, especially since the primary MX is available to do the
> callout (to avoid shedloads of spam emails getting stuck in the
> secondary MX).

Maybe something to do with the exim que run which runs every 30 minutes?
Either force immediate delivery from the secondary or reduce the
QUEUEINTERVAL in /etc/default/exim4.

Does your secondary MX simply spool mails and then forward them to your
primary so if your primary goes down they are held on the secondary
until it comes back up?  In theory, the sending hosts will resend anyway
if the primary goes down - in which case, when you fix the primary
they will succeed. I would only use a secondary if the volume was such
that one server couldn't cope and, thankfully, I've never reached that
point. Anything to avoid administering unnecessary mailservers :)

Cheers

John