Re: routing table problem after power failure (sarge)
From: Jan C. Nordholz (_at_murphy.debian.org)
Date: Mon, 24 Jan 2005 21:42:20 +0100 To: email@example.com
On Mon, Jan 24, 2005 at 05:41:19PM +0100, Corné wrote:
> Hi. I've got this problem:
> - Running Debian Sarge 2.4 for a month succesfully.
> - After playing around with Postfix' configuration, I can't connect to the
> internet any more. Power failure somewhere during config too.
> - static ips, connected through internet by router/modem
> - two interfaces, eth0 and eth1, both on lan (router on lan too)
> - router is known to be working (I send this mail ;-) )
> - was desperate, so I installed sendmail after this shit, still in apt-get
> Samba is working on my lan, but I can't even ping my ISP's dns server
> (package loss 100%). Pinging to ip adresses on my lan, including router, is
> ok. Pinging 127.0.0.1 is ok, but pinging localhost times out.
You should add localhost to /etc/hosts so that your host doesn't try to
resolve that name (which it seems to do, and fails):
(example of your /etc/hosts)
10.0.0.1 (name of your box)
10.0.0.2 (name of your box)
10.0.0.250 (router's name)
(end of example)
> After every reboot, /etc/resolv.conf is clean: all content removed. I put
> back "nameserver x.x.x.x" and save.
Hmm, don't know what tool would dare to wipe resolv.conf... anyway, putting
back one or more nameserver entries doesn't get DNS going? Try to call
the nameserver manually: use a tool like "dig" (contained in the package
dnsutils) oder "host" (in its own package) to issue a query, and see
if you get a response. There may be a problem with the nameservers you
> I can't use http and ftp to the box any more, but I can SSH.
At least http _requires_ hostname resolution - I'm not sure about ftp.
SSH will work fine with IPs only.
> [snip ifconfig output]
> I think this is ok (?).
Yep, seems ok. But one question: Why do you have two interfaces
connected to the same network? Are they firewalled differently?
> route -n returns this. i think this is wrong:
> Kernel IP routing table
> Destination Gateway Genmask Flags Metric Ref Use
> 10.0.0.0 0.0.0.0 255.0.0.0 U 0 0 0 eth0
> 10.0.0.0 0.0.0.0 255.0.0.0 U 0 0 0 eth1
> 0.0.0.0 10.0.0.250 0.0.0.0 UG 0 0 0 eth1
> 0.0.0.0 10.0.0.250 0.0.0.0 UG 0 0 0 eth0
Well, that's basically ok, I think - two default routes via your router,
and two LAN routes. All metrics are 0, but route should get along with
that. ;-) And if you can still send mail and ask the list, at least
this part is working.
> The box is not working as an internet sharing device. bind9 is installed,
> don't know if it has been working before.
Bind is a tricky thing to configure if you've never seen DNS zone description
files before - is there any special reason you're running bind?
You can easily check if bind's working: issue a DNS query aimed at 127.0.0.1.
> The main issue: nslookups not working
> smaller issue: localhost not found.
> Can somebody give me some hints how to repair this, without a clean install?
First step: Try DNS manually on the servers you've got
Second: If you get no response, ping them...
Third: If all this fails or if you're not sure about the nameserver IPs you've got,
check if one of the TLD servers is responding to you - for example B.NIC.FR
(22.214.171.124), one of the authoritative servers for .nl domainnames.
Finally: Report back your experience. :-)
> Thanx, Corné (Holland)
-- Jan C. Nordholz <jckn At gmx net>
-- To UNSUBSCRIBE, email to debian-user-REQUEST@lists.debian.org with a subject of "unsubscribe". Trouble? Contact firstname.lastname@example.org
- application/pgp-signature attachment: Digital signature