Skip site navigation (1)Skip section navigation (2)
Date:      Mon, 6 Jan 1997 15:00:17 -0600 (CST)
From:      Jimbo Bahooli <moke@fools.ecpnet.com>
To:        Giles Lean <giles@nemeton.com.au>
Cc:        freebsd-security@freebsd.org
Subject:   Re: sendmail....tricks... 
Message-ID:  <Pine.BSF.3.95.970106144729.277A-100000@fools.ecpnet.com>
In-Reply-To: <199701060904.UAA00711@nemeton.com.au>

next in thread | previous in thread | raw e-mail | index | archive | help
On Mon, 6 Jan 1997, Giles Lean wrote:

> 
> On Sun, 5 Jan 1997 18:47:29 -0600 (CST)  Jimbo Bahooli wrote:
> 
> > The first idea, which i have successfully accomplished, is logging and
> > access control via tcp wrappers.
> 
> Interesting; I think I'd go about it differently:
> 
> Since sendmail currently supports using libwrap from Wietse Venema's
> tcp_wrappers distribution, this could be used to block non-local
> access to sendmail.  With remote access to sendmail blocked it can use
> a non-standard port and smap/smapd from the TIS firewall toolkit could
> be used to talk to strangers.
> 
> (Alternative to libwrap is one of the in-kernel firewalling solutions,
> but I don't think these log as well as application level checking, and
> must lose at least a little in performance for ordinary traffic.)

Going into the experiment I was just trying to transparently move sendmail
to a different port, the logging and access control came about from the
use of tcp wrappers from inetd.  I figured this a plus and decided to add
that in. When time permits I am going to work on moving it to a non-root
port and sendmail will run soley as user mailer.

Another idea, but since I do not know the excacts of sendmail, would be to
run a program to bind to port 25. Then start sendmail as user mailer or
some other person. I understand this can be done from inetd, but a new
sendmail is started each session which is alot of excess overhead even on
systems that do not pass much mail.






Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?Pine.BSF.3.95.970106144729.277A-100000>