ping: sendto: host is down
Christoph Leuzinger
I got a strange networking problem. I'm currently running NetBSD
1.4.2 on i386 with a 3c509 in it.
I did the installation via FTP from a host in the LAN and everything
went well. But as I now boot for the first time, I just can't get any
network access. (I'm using just the same network config as sysinst,
though.) I can't ping any host in the LAN, nor can I ping the
problematic host from outside.
Another strange thing is that ping is _very_ slow. It takes about
half a minute or sometimes even longer 'til I get a response from
"ping 127.0.0.1".
I've checked half the /etc so far, but still can't find the mistake.
Any hints? TIA.
Regards,
Chris.
--
__________________________________________________________
Christoph F. Leuzinger + Student @ Kantonsschule
http://www.westworks.ch/ + Schaffhausen, Switzerland
ICQ# 20603375 + http://www.kanti.ch/
Jason R. Fink
ali (Anders Lindgren)
> Christoph,
>
> Have you tried just using ifconfig?
I have an 1.4.2/i386 box (486dx2/66) at home. After the upgrade
to 1.4.2, I seem unable to get networking to work. I can't even
ping my gateway. Ifconfig:ing the interfaces and pinging the
interfaces' own IPs works fine. Everything seems to work, but
the network's just gone.
Tonight I succesfully[0] installed NetBSD-1.5_ALPHA2 on my Amiga
to get my BSD fix while using that one too. It was previously
running a quite well working port-amiga/1.4.1-GENERIC kernel.
After the install and configuration, I get the same phenomenon
on the amiga! I ran tcpdump and got lots of broadcast traffic,
but ping/traceroute/<insert tool here> won't work. It won't
send! IP-numbers were used while pinging and tracerouting, so
it shouldn't be due to the (not yet configured) DNS stuff.
At first I thought my i386 problem was due to my compiling a
custom kernel (for my two NE2k ISA NICs) with ipf/ipnat enabled,
but it doesn't seem to be connected to ipnat/ipf (because it
doesn't work after a reboot disabling those either).
Did something seriously break somwhere in 1.4.2?
On the amiga side, I get the usual "le0: lost carrier" message
I've gotten on it since 1.3.3 (it says so when the Ariadne card
is first configured but it works: used to anyway). This does
not happen on the ne0 and ne1 interfaces of the i386 box.
Sorry for the cross-posting, but this is exactly what I am
experiencing on both my i386 and now also on my newly installed
1.5_ALPHA2 amiga.
/ali - distressed! ;(
Manuel Bouyer
> Hello,
>
> I got a strange networking problem. I'm currently running NetBSD
> 1.4.2 on i386 with a 3c509 in it.
>
> I did the installation via FTP from a host in the LAN and everything
> went well. But as I now boot for the first time, I just can't get any
> network access. (I'm using just the same network config as sysinst,
> though.) I can't ping any host in the LAN, nor can I ping the
> problematic host from outside.
What does 'ifconfig -a' and 'netstat -rn' report ?
>
> Another strange thing is that ping is _very_ slow. It takes about
> half a minute or sometimes even longer 'til I get a response from
> "ping 127.0.0.1".
May be related to the first problem, name resolution times out.
--
Manuel Bouyer <bou...@antioche.eu.org>
--
Jason R. Fink
I have seen this somewhere else, I lost the original post :(
I am going to fire off some quickies about stuff that has happened
to me:
- I once had a wake up on LAN card (with NetBSD/i386-1.4.2)
that got so frustrating I ended up putting another card in
- the card may be a 10/100 which may need a different media type
I once had an autoselct 10/100 that if i tried to say
just 10 or just 100 in wouldn't work, I think I actually
called the media 'autoselect' - or somehting, I cannot
remember
email me back the card type again and i will check tomorrow
at some point if it is the same,
regards,
jrf
--
Jason R. Fink <j...@diverge.org>
ali (Anders Lindgren)
[snip stuff about network not working although the interfaces do]
I solved my problems, both on the 1.5_ALPHA2 amiga and the
1.4.2 i386 firewall that I am currently using:
The problem with my amiga was that I had typed in the wrong
IP-address to my DNS. Apparently, NetBSD network utilities
(ping, traceroute...) will attempt to contact the DNS even
if supplied with an IP-address[0]. When I fixed the bad DNS
address everything automagically started working just fine!
On the i386 it was a completely different problem (the DNS
addresses were correct): An EPROM option of the external-link
NIC that doesn't show up with ifconfig et al was wrong, and
when I changed it, everything worked fine.
I now have a neatly working little intranet, sharing my
single real-world IP through the NATing i386 box, and
I'm happier than ever with this fantastic OS. :-)
In short, they were all EBSAC-errors[1]. ;O)
[0] Maybe this should be documented? Pinging an IP-address
will seem to fail if the DNS is unreachable, which is
very confusing (IMHO).
[1] Error Between Screen And Chair. ;o)
--
/ali: Computer Science Major and aspiring cartoonist. :-)
(dept) dat9...@ludat.lth.se - http://www.ludat.lth.se/~dat94ali
(home) a...@h543.sparta.lu.se - http://h543.sparta.lu.se/
* A4000/040-40/CV3D/Ariadne·AmigaOS·NetBSD·A3000/040-25/Ariadne *
David Brownlee
> The problem with my amiga was that I had typed in the wrong
> IP-address to my DNS. Apparently, NetBSD network utilities
> (ping, traceroute...) will attempt to contact the DNS even
> if supplied with an IP-address[0]. When I fixed the bad DNS
> address everything automagically started working just fine!
>
Both ping and traceroute document the -n option:
-n Numeric output only. No attempt will be made to look up symbolic
names for host addresses.
> I now have a neatly working little intranet, sharing my
> single real-world IP through the NATing i386 box, and
> I'm happier than ever with this fantastic OS. :-)
>
Excellent to hear :)
> [0] Maybe this should be documented? Pinging an IP-address
> will seem to fail if the DNS is unreachable, which is
> very confusing (IMHO).
What would you suggest - a CAVEATS section at the end of the
ping man page?
David/absolute
-- www.netbsd.org: A pmap for every occasion --
Jason R. Fink
| ping man page?
I think it is worth mentioning (for all) that the NetBSD Guide
http://www.mclink.it/personal/MG2508/nbsdeng/netbsd.html
covers this scenario pretty well. Perhaps I should have suggested
that right off the bat :)
--
Jason R Fink <j...@diverge.org> WWWhome <www.diverge.org/jrf>
"Upon completion of maintenance discard spare parts and leftover screws"
Christoph Leuzinger
> > network access. (I'm using just the same network config as sysinst,
> > though.) I can't ping any host in the LAN, nor can I ping the
> > problematic host from outside.
> What does 'ifconfig -a' and 'netstat -rn' report ?
# ifconfig -a ep0:
flags=8863<UP,BROADCAST,NOTRAILERS,RUNNING,SIMPLEX,MULTICAST> mtu 1500
address: 00:20:af:35:51:d0 media: Ethernet 10base2 inet
192.168.0.2 netmask 0xffffff00 broadcast 192.168.0.255 lo0:
flags=8009<UP,LOOPBACK,MULTICAST> mtu 32976 inet 127.0.0.1 netmask
0xff000000 ppp0: flags=8010<POINTOPOINT,MULTICAST> mtu 1500 ppp1:
flags=8010<POINTOPOINT,MULTICAST> mtu 1500 sl0:
flags=c010<POINTOPOINT,LINK2,MULTICAST> mtu 296 sl1:
flags=c010<POINTOPOINT,LINK2,MULTICAST> mtu 296 strip0: flags=0<> mtu 1100
strip1: flags=0<> mtu 1100 tun0: flags=10<POINTOPOINT> mtu 1500 tun1:
flags=10<POINTOPOINT> mtu 1500 gre0: flags=8010<POINTOPOINT,MULTICAST> mtu 1450
gre1: flags=8010<POINTOPOINT,MULTICAST> mtu 1450 ipip0:
flags=8010<POINTOPOINT,MULTICAST> ipip1: flags=8010<POINTOPOINT,MULTICAST>
eon0: flags=3<UP,BROADCAST> mtu 1500
# netstat -rn
Routing tables
Internet:
Destination Gateway Flags Refs Use Mtu Interface
default 192.168.0.1 UGS 0 8 - ep0
127.0.0.1 127.0.0.1 UH 1 30 - lo0
192.168 link#1 UC 0 0 - ep0
192.168.0.1 link#1 UHL 1 5 - ep0
XNS:
Destination Gateway Flags Refs Use Mtu Interface
ISO:
Destination Gateway Flags Refs Use Mtu Interface
X.25:
Destination Gateway Flags Refs Use Mtu Interface
AppleTalk:
Destination Gateway Flags Refs Use Mtu Interface
> > Another strange thing is that ping is _very_ slow. It takes about
> > half a minute or sometimes even longer 'til I get a response from
> > "ping 127.0.0.1".
> May be related to the first problem, name resolution times out.
# ping -n 127.0.0.1
This works and now it's fast. But only localhost.
I can't see the error, there are differences between the outputs of ifconfig
and netstat on the problematic host and another NetBSD1.4.2 machine here (which
runs fine, with another NIC in it).
Greets,
Lynn S. Weeks
you have a new card installed - you didn't plug the new card into the
uplink port did you? This baffled and confused me the first time I did
it. My brother made enough fun of me that I realized exactly what I had
done the second time.
Christoph Leuzinger
I got my problem resolved now. It was because NetBSD doesn't accept the 3c509
if it's not set to IRQ10 (it's detected, though, but wouldn't run). So I simply
changed it and everything goes allright now.
Thanx to everybody here and in #netbsd!
Greetings, Chris.