Date: Thu, 11 Jan 1996 02:40:40 +0000 From: davidn@unique.usn.blaze.net.au (David Nugent) To: rif@ns.kconline.com (Jim Riffle) Cc: freebsd-isp@FreeBSD.org Subject: Re: MS Exchange client Message-ID: <Mutt.19960111024040.davidn@labs.blaze.net.au> In-Reply-To: <Pine.BSI.3.95.970111001406.10148A-100000@rif.kconline.com>; from Jim Riffle on Jan 11, 1997 00:34:24 -0500 References: <9701110453.AA28334@wr.com.au> <Pine.BSI.3.95.970111001406.10148A-100000@rif.kconline.com>
next in thread | previous in thread | raw e-mail | index | archive | help
Jim Riffle writes: > > rif> So, the problem lies in having my box accept the mail for their domain and > > rif> hold it in the queue until it is deliverable. > > > > See Sendmail book and DNS book(For detail, see http://www.ora.com) > > Thank you for the suggestion, I just may have to break down and actually > get that book sometime. I bet that would have made setting up UUCP a lot > eaiser when I got that one going. I'll break it down and make it easy for you. It isn't all that complex. It is a good idea to get the sendmail guide regardless. > The weird thing which I just don't understand on how this should work is > if I give a MX record for say foo.com so that mail is handled by > 207.51.167.3, and then on 207.51.167.3, put in a CW line so that it will > accept mail for the domain, how in the world is it ever going to get to > the client who dials up saying they are foo.com. You don't want the cw line. This is only for aliases of YOUR host, which is not what you want to do. You need to requeue all their mail for their smtp daemon and ignore the DNS. > to have 207.51.167.3 handle its mail, so when sendmail processes its que, > it would seem that it would try to send the mail right back to itself. I recall reading about a new control line in sendmail which you can tell it to deliver "direct" if you're the final MX target for a domain/host and it isn't a local name. I don't recall what it is, but that is probably the easiest method. I use something slightly different. > Is this a simple issue of having 2 MX records, having their machine having > a higher priority? No. Then you're no longer the relay, but the fallback MX. The method I'll explain involves simply adding explicit routing for the host on your system. Basically, you want sendmail to route specially for that domain, and not use the MX. As I said earlier, apparently 8.8.x has a new keyword which can make this semi-automatic (add the keyword and you're done), but the following is what I've been using for years. Go back to the .mc (m4 macro version of the cf file) and add the line: FEATURE(mailertable, 'hash -o /etc/mailertable')dnl Generate the .cf and installed it. Create the file /etc/mailertable and add: .your.clients.domain smtp:[host.your.clients.domain] Run: makemap hash /etc/mailertable which will generate the .db version in /etc/mailertable.db. Restart sendmail. QED. /etc/mailertable is just a local routing configuration table that tells sendmail to map a given host or domain (domain entries must start with a '.') to a given mailer. Using the square brackets around the host to queue for overrides use of the MX, otherwise you can create an MX loop. The above will queue all mail to *.your.clients.domain to host.your.clients.domain. Ah - found it. I've never tried it, but the variable that does this automatically in sendmail 8.8.x is called 'TryNullMXList'. Quoting cf/README from the sendmail sources: confTRY_NULL_MX_LIST TryNullMXList [False] If we are the best MX for a host and haven't made other arrangements, try connecting to the host directly; normally this would be a config error. Regards, David Nugent - Unique Computing Pty Ltd - Melbourne, Australia Voice +61-3-9791-9547 Data/BBS +61-3-9792-3507 3:632/348@fidonet davidn@freebsd.org davidn@blaze.net.au http://www.blaze.net.au/~davidn/
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?Mutt.19960111024040.davidn>