[Exim] host_accept_relay problem with netscape on localhost

Página superior
Eliminar este mensaje
Responder a este mensaje
Autor: Daniel Einspanjer
Fecha:  
A: exim-users
Asunto: [Exim] host_accept_relay problem with netscape on localhost
I have been working for about a month on getting Exim set up with my virtual
domains including SMTP auth and having a POP3 server to let people access
it. (Yes it's way to long but I'm a linux newbie. :) I've got the auth
working and everything but I currently have one problem that I am lost with:

Netscape Communicator on my local linux box (the one running exim) always
requires authentication. I had a horrible time getting host_accept_relay to
work in the first place since I originally wanted it to inlude just local
host and the four IP addresses in my local network. I did some fishing
around but I couldn't seem to find a mask that would work for that.
Currently I am using the mask 208.36.207.0/24 which as I understand it,
allows anything on that class C to relay.. :/ I tried two different methods
of file listing:

host_accept_relay = lsearch;/etc/virtual/localips    and
host_accept_relay = /etc/virtual/localips
where localips was a simple text listing of each IP address.



I noticed that when netscape sends mail, it uses the IP localhost so I added
localhost to my list: host_accept_relay = localhost:208.36.207.0/24 but
netscape still prompts me for a password whenever I try to send mail. I saw
a couple of FAQs about this and someone had posted a patch to exim to make
it not offer the AUTH command to someone who matched host_accept_relay, but
when I went to look at the code, that patch was already there (I'm running
exim 3.13) and it doesn't seem to work.. ??

One thing I was wondering is that when I do -bh tests, host_accept_relay is
not checked until I list an RCPT that is not in local_domains. I was
wondering how this is supposed to work since if a client issues an EHLO
command, exim automatically returns with AUTH as one of the available
commands. I did not see anything in the logging about host_accept_relay
being checked at that time.

Could anyone shed some light on this problem for me? Am I missing a setting
somewhere that causes host_accept_relay to be checked earlier?

Thanks,

Daniel Einspanjer
mailto:daniel@yipyip.com