On 4 Jul 1999, Lukasz Grochal wrote:
> 1) There is a subnet (say - 192.168.0.0/24), that I want to relay all
> mail for. As those are generally trusted clients (company's employees
> on an intranet network), I'd like to accept any envelope sender and
> 'From: ' field, no matter what they contain (erm, virtually, standard
> validity checks apply ;).
host_accept_relay = 192.168.0.0/24
> 2) Because of a very weird situation in the dial-up market here in
> Poland, I also have to relay mail from other hosts (dialup modems,
> say *.blah.blurp.pl, various subnets, practically no aggregate info
> on their IP addresses available, only the reverse DNS lookups
> can be used to find who they are, hence the *.blah.blurp.pl pattern)
> and I want those (and only those, except of the subnet mentioned
> above) clients to be able to use my server as relay if and only if
> they have valid_user@??? in envelope / 'From:' fields.
host_accept_relay = 192.168.0.0/24 : *.blah.blur.pl
You can set sender_address_relay, but it applies to all messages, not
just those from a specific set of hosts. Unfortunately the value of
sender_address relay is not string-expanded, so there is no way you can
vary it depending on the host. It would probably be a useful addition to
make it an expanded string, so I have added this idea to the Wish list.
> 3) As I already mentioned above, no other hosts should have any rights
> to use my server as a relay, even if they introduce a valid
> user@??? in 'MAIL FROM:' command.
That would be the case, provided you do NOT set relay_match_host_or_sender.
> 4) To make things even more complicated (?, it shouldn't actually make
> much difference, but I want to make things as clear as possible),
> the mail server is a primary MX for a set of domains (*.rmf.pl,
> *.rmf.fm, etc. etc.)
Those should be in local_domains or relay_domains, as appropriate.
> a) There would be no problem if I denied users from 192.168.0.0/24
> use the server as a relay if sender address is not a local one.
> But I really want to avoid this.
I don't think there's an easy way at the moment.
--
Philip Hazel University of Cambridge Computing Service,
ph10@??? Cambridge, England. Phone: +44 1223 334714.
--
*** Exim information can be found at
http://www.exim.org/ ***
