[snip]
I think the problem is Postfix is sending using eth0, which in turn
means that it appears to come from n.n.n.186, which in turn means that a
reverse lookup does not resolve to mail.domain.tld. The loop is not
closed and therefor we are suspect.
then why not fix that?
I would like to, unfortunately I cannot do this at the moment some
custom client software uses fixed IP addresses rather than doing the
right thing and using DNS. fortunately we have decided to drop it in 3 -
6 months.
n.n.n.186 PTR foo.example.com
n.n.n.187 PTR bar.example.com
foo.example.com A n.n.n.186
bar.example.com A n.n.n.187
if that's not possible, then maybe
n.n.n.186 PTR mail.example.com
n.n.n.187 PTR mail.example.com
mail.example.com A n.n.n.186
mail.example.com A n.n.n.187
[snip]
to define the IP used to send mail, use smtp_bind_address. however, if
you have mail that should go through another interface, then you'd
better define two smtp transports and configure each with its
smtp_bind_address (using -o), then setup transport entries to select
which smtp to use.
No, all mail goes through the one interface, so I added the
smtp_bind_address to master.cf, and it seems to work like a charm.
It has not solved the problem with AT&T, but I am not sure what the real
problem is. I have written to the Postmaster outlining the problem we
are having, now I just have to wait for a reply. I am not holding my breath.
do not touch inet_interfaces unless you really need that. your example
(inet_interfaces=ip1 ip2 localhost) looks useless.
I thought so to. I could not see the difference between the default
value of all and the suggest value of ip1 ip2 localhost, as a
far as I could see the default value would be ip1 ip2 localhost.
Thanks for the input
John A
--
"All that is necessary for the triumph of evil is that good men do nothing."
(Edmund Burke)
