Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

Sendmail 8.9.1

1 view
Skip to first unread message

Dan Davis

unread,
Mar 16, 1999, 3:00:00 AM3/16/99
to
Can anyone point me to a useful site concerning configuration of this
version? In particular I need to update my cf file from a very old
installation.

Thanks,
Dan Davis


Ben Gibbs

unread,
Mar 18, 1999, 3:00:00 AM3/18/99
to
Apart from the obvious www.sendmail.org I heartily recommend you go to a
bookstore and buy the O'Reilly book "sendmail" also known as the batbook.
This will tell you everything. Also, download the latest source and use the
m4 compiler to generate your .cf file, it's much easier and safer than
hacking sendmail.cf IMO.

Ben

Dan Davis wrote in message <36EEEB37...@mprg.ee.vt.edu>...

Per Hedeland

unread,
Mar 18, 1999, 3:00:00 AM3/18/99
to
In article <36EEF03E...@vt.edu> Dan Davis <dada...@vt.edu> writes:
>============ SYSTEM IDENTITY (after readcf) ============
> (short domain name) $w = solarsys
> (canonical domain name) $j = solarsys.ee.vt.edu
> (subdomain name) $m = ee.vt.edu
> (node name) $k = solarsys
> (mailserver) ${ms} = sol
>========================================================
>
>Where does sendmail get this info from? Particularly (node name) and
>(subdomain name).

Sendmail 8.9.1 gets it through starting with gethostname(), and if that
doesn't contain at least one '.', hunting through the name services
specified for "hosts" trying to find a name that does. Assuming it finds
one, it uses the part before the first '.' for (node name) (and for
(short domain name)), and the part after it for (subdomain name).

However, you're not running 8.9.1 but 8.9.1b+Sun, which is 8.9.1 with
Sun's private mods: 8.9.1 from sendmail.org will never find a
(mailserver) - that's the host /var/mail is mounted from, and of course
sendmail has no business knowing that /var/mail is used for anything in
particular. I believe 8.9.1b+Sun throws the NIS and/or NIS+ domain name
into the algorithm above for, um, "dubious" reasons.

--Per Hedeland
p...@erix.ericsson.se

0 new messages