Petre, yes, my /etc/resolv.conf file is set up correctly. Thanks. Lars,
> Charles Bray wrote: > > > Right now, my /etc/network/interfaces file includes the following > lines > > (excluding comments): > > > > auto lo > > iface lo inet loopback > > > > My understanding is that the "auto lo" line is supposed to activate > the > > loopback interface at boot time. But evidently it doesn't. After I > > > boot and log in as root, if I ping the localhost at 127.0.0.1, I get > the > > error message: > > > > neighbor table overflow > > > > At this point, the routing table (route -n command) doesn't list the > > > loopback interface, and I must manually activate and add it with the > > > lines: > > > > ifconfig lo 127.0.0.1 > > route add 127.0.0.1 lo > > The route looks wrong to me. Something like > > route add -net 127.0.0.0 netmask 255.0.0.0 lo > > should be what you need.. So the lo device will take all traffic on the entire 127.0.0.0 network, not just at 127.0.0.1. Is this preferred? > Instead of the two command you might want to try running > > ifup lo I can't tell that this command does anything, but the "route add -net.... " line works fine, although it doesn't seem to be needed (see below). > > Strangely (to me), if I comment out the "auto lo" line in > > /etc/network/interfaces and reboot, I can then successfully ping the > > > loopback address, indicating that the interface has been > automatically > > activated. So, as of now, I've commented out the "auto lo" line to avoid the above-described "neighbor table overflow" error, and, although the loopback interface isn't listed in the routing table, everything works just fine. I can ping localhost or 127.0.0.x with no problem. I wish I understood, but I guess I don't have to right now. > > After successfully connecting to my ISP via serial modem (and after > > manually adding the loopback interface), the routing table is as > > follows: > > > > Destination Gateway Genmask Flags Metric Ref Use > Iface > > 286.115.220.139 0.0.0.0 255.255.255.255 UH 0 0 0 > ppp0 > ---^ Something is definitely very very wrong here. Each field of the > ip > address should be in the interval 0..255, 286 just cannot happen... Of course, you're right. I'm sorry. I tried to be very careful not to make any mistakes, but the 8 in 286 was actually one of those zeros with a diagonal line through it, and it looked like an 8 to my weary eyes. Since my Linux computer isn't ready yet, I'm posting from my wife's Windows machine. > > 127.0.0.1 0.0.0.0 255.255.255.255 UH 0 0 0 lo > > 192.168.1.0 0.0.0.0 255.255.255.0 U 0 0 0 > eth0 > > 0.0.0.0 192.168.1.1 0.0.0.0 UG 0 0 0 > eth0 > > > > I can ping 286.115.220.139 as well as my machine's local network IP > > address at 192.168.1.1, but I can't ping my ISP's DNS servers or > > anywhere else on the internet. > > If you can ping that illegal address something is completely utterly > wrong. You make me laugh. :) > Did you copy the info by hand and make a typo ? < : ) > Your routing table looks very weird to me. Lets have a look... > First of you need the loopback route: > > 127.0.0.0 0.0.0.0 255.0.0.0 U 0 0 > 0 > lo > > If its not present you should run > route add -net 127.0.0.0 netmask 255.0.0.0 lo Again, this doesn't seem to be necessary, but even when I do run this command, it's not permanent. After reboot, it's gone. How can I fix this? > Your box is capable of finding itself without a route I'm not sure I understand this. > so the strange 286.... route should go. Experiment with "route del" to > get rid of it. The "286" route was designated automatically when I dialed into my ISP. It's the ISP's PPP server address. > When looking for other addresses you need to first of all be able to > find your gateway. ......... Please read my reply to Michael Heldebrandt regarding the gateway problem. Thanks for all your advice and time. I deeply appreciate your help. Charles
