Ping is broken after Kernel upgrade

wolverene13

I upgraded my Linux kernel (Debian Lenny --> Squeeze) a few weeks ago and now ping is broken...I don't see failed pings like I used to. Normally I would see "Destination host unreachable," but now it shows nothing in real time until I Ctrl+C out of it:

allen@debian:~$ ping 1.1.1.1
PING 1.1.1.1 (1.1.1.1) 56(84) bytes of data.
^C
--- 1.1.1.1 ping statistics ---
10 packets transmitted, 0 received, 100% packet loss, time 9021ms

allen@debian:~$

See? It shows that 10 packets failed, but there's no real-time output like there used to be. It's not a huge deal, but I would like it to work as designed.

Any ideas?

krauser

I did it too, but it went uglier than your condition. I couldn't even boot into the system, the screen went blank. It's a risky attempt (50-50). I rolled back for solving.

Met44

Host unreachable messages should be generated when an ARP request times out somewhere. Try pinging an address on your local subnet that isn't assigned and see if you get the message.

Forsaken_GA

wolverene13 wrote: »

I upgraded my Linux kernel (Debian Lenny --> Squeeze) a few weeks ago and now ping is broken...I don't see failed pings like I used to. Normally I would see "Destination host unreachable," but now it shows nothing in real time until I Ctrl+C out of it:

allen@debian:~$ ping 1.1.1.1
PING 1.1.1.1 (1.1.1.1) 56(84) bytes of data.
^C
--- 1.1.1.1 ping statistics ---
10 packets transmitted, 0 received, 100% packet loss, time 9021ms

allen@debian:~$

See? It shows that 10 packets failed, but there's no real-time output like there used to be. It's not a huge deal, but I would like it to work as designed.

Any ideas?

I'm actually used to the behavior you're describing, as all of lenny boxen and my squeeze box do the same thing. So I wouldn't say ping is broken, it's just not operating as you'd expect. Now that I think about it, I think my mac is the only one that shows me destination unreachable.

if the only thing that's changed is the kernel (ie, you didn't update anything like the iputils-ping package), it's probably just a kernel variable that you need to turn back on

demonfurbie

have ya tried pinging 127.0.0.1 and or have ya tried arping?

Forsaken_GA

Met44 wrote: »

Host unreachable messages should be generated when an ARP request times out somewhere. Try pinging an address on your local subnet that isn't assigned and see if you get the message.

You know what, I feel absolutely retarded, that makes perfect sense, and you are indeed correct.


forsaken@winterfell:~$ ping 192.168.12.253
PING 192.168.12.253 (192.168.12.253) 56(84) bytes of data.
From 192.168.12.239 icmp_seq=1 Destination Host Unreachable
From 192.168.12.239 icmp_seq=2 Destination Host Unreachable
From 192.168.12.239 icmp_seq=3 Destination Host Unreachable
^C
--- 192.168.12.253 ping statistics ---
5 packets transmitted, 0 received, +3 errors, 100% packet loss, time 4007ms

forsaken@winterfell:~$ ping 1.1.1.1
PING 1.1.1.1 (1.1.1.1) 56(84) bytes of data.
^C
--- 1.1.1.1 ping statistics ---
34 packets transmitted, 0 received, 100% packet loss, time 33014ms

Since the local stack knows the offnet address is offnet, it won't issue an arp request for it and just shunts it off to the default gateway.

I can see I need to review my TCP/IP basics.