On 13.06.2017 17:34, Bill Pye wrote:
>
> I've got most of the configuration corrected, I believe, but there's still one thing that's eluding me. Do I need to define a separate unix socket for the milter process (I have configured it correctly in the MTA) in 1.6 and if so, where, or do I modify the "control_socket"? I can't actually see anything in the documentation that actually describes that process unless I'm missing something blindingly obvious. Unfortunately I'm the sort of person that prefers a 'cook book' type of description with worked examples rather than the more technical details currently on the rspamd web site, that's not a complaint it's just the way I work. :)
>
Bill, consider using TCP socket as it is enabled in Rspamd 1.6 out of the box. Using Unix socket is a little bit complicated, at least you need to set permissions and remove socket file by yourself. AFAIK there were problems with Unix sockets on some OSes, systemd-based in particular.
In self-scan mode Rspamd integration with Postfix is simple as:
Postfix
main.cf:
smtpd_milters = inet:localhost:11332
milter_mail_macros = i {mail_addr} {client_addr} {client_name} {auth_authen}
milter_protocol = 6
milter_rcpt_macros = i {rcpt_addr}
# skip mail without checks if milter will die
milter_default_action = accept
/etc/rspamd/local.d/worker-proxy.inc:
upstream "local" {
self_scan = true;
}
To enable extended headers in messages you need just one line:
/etc/rspamd/local.d/milter_headers.conf:
extended_spam_headers = true;