Return to BSD News archive
Newsgroups: comp.unix.bsd.freebsd.misc
Path: euryale.cc.adfa.oz.au!newshost.carno.net.au!harbinger.cc.monash.edu.au!munnari.OZ.AU!news.Hawaii.Edu!news.uoregon.edu!arclight.uoregon.edu!nntp.primenet.com!howland.erols.net!news.mathworks.com!uunet!in1.uu.net!sys14!news
From: Con Sultant <jkane@execpc.com>
Subject: Boot hangs with resolv.conf available.
X-Nntp-Posting-Host: pc105164.jdt.deere.com
Content-Type: text/plain; charset=us-ascii
Message-ID: <3277AD06.1CAB@execpc.com>
Sender: news@deere.com
Reply-To: jkane@execpc.com
Content-Transfer-Encoding: 7bit
Organization: Yes I Am.
Mime-Version: 1.0
Date: Wed, 30 Oct 1996 19:31:18 GMT
X-Mailer: Mozilla 3.0 (Win16; I)
Lines: 49
I finally got around to figuring why my machine hangs at boot time. I
have the problem where it hung after ifconfig. It is caused by the
combination of using a dns resolver in /etc/resolv.conf that is not on
the local net. And having a dial on demand solution for after boot
time.
After seeing the many messages here from others in the same situation, I
took the advise of fixed IP address to heart. I went into the
/etc/sysconfig file, and added a new variable called $ip_address. I
substituted the ifconfig line ip address with this variable, and set the
variable to my ip address. I then looked at the routing portion. That
was where the hang occured. I found that the $route_loopback hung no
matter what since it had already been set (from where I do not know) but
it was trying to resolve, and then couldn't write to the socket, since
there was already an entry for it! I remarked it out. I then removed
loopback from the $static_routes variable as included in the
distribution. I changed the $route_multicast line to replace the
${hostname} with my new $ip_address. This prevented it from doing a
domain lookup on the $hostname. Why was this like this in
distribution????
That has gotten me much farther. Now, It still hangs on sendmail.
Presumably doing a DNS lookup or something. This is stock
configuration. Is there something in the /etc/sendmail.cf file or a
command line option to sendmail that will stop it from doing a lookup at
boot time, and waiting till it is running for say 30 minutes or
something. (Yes, I meant to say 30 minutes since that is the normal
interval.)
I killed sendmail, and restarted it with the switches install put in
/etc/sysconfig, (-bd -q30m) and it dialed and sent a couple of packets.
This confirmed to me that something is causing it to get to another
network. Since it doesn't do this without a resolv.conf, I figure this
must be a DNS lookup.
Please at least tell me if I am looking in the right direction. Better
yet, What in sendmail.cf needs to be changed? I am not doing any
mailing outside of the box itself, but do send mail to myself and others
on this machine. I may some day want to use it as a mail pop server
since it does have internet access.
Thanks for any guidance.
--
Jeff Kane mailto:jkane@execpc.com
Sysop - Home Brew University BBS Brew City Campus 414-238-9074
Genealogy Search: Brickner, Kane, Kimbro, LaClaire, Snyder|Snider,
Seecs|Seetch|Sich, and Thorton.
Visit my homepage for more ... http://www.execpc.com/~jkane