Linux - NetworkingThis forum is for any issue related to networks or networking.
Routing, network cards, OSI, etc. Anything is fair game.
Notices
Welcome to LinuxQuestions.org, a friendly and active Linux Community.
You are currently viewing LQ as a guest. By joining our community you will have the ability to post topics, receive our newsletter, use the advanced search, subscribe to threads and access many other special features. Registration is quick, simple and absolutely free. Join our community today!
Note that registered members see fewer ads, and ContentLink is completely disabled once you log in.
If you have any problems with the registration process or your account login, please contact us. If you need to reset your password, click here.
Having a problem logging in? Please visit this page to clear all LQ-related cookies.
Get a virtual cloud desktop with the Linux distro that you want in less than five minutes with Shells! With over 10 pre-installed distros to choose from, the worry-free installation life is here! Whether you are a digital nomad or just looking for flexibility, Shells can put your Linux machine on the device that you want to use.
Exclusive for LQ members, get up to 45% off per month. Click here for more info.
I have read other posts with this problem. None of the solutions worked for me. I have a total of three computers, one Windows XP, one Fedora 5, and one I installed with Fedora 5 but can't get networking to start. The funny thing is that the Fedora 5 that works is the same hardware (emachine T3304) with the same amount of memory, same disc space, as the machine doesn't work. LVM configured a little different and a second disc has Windows XP on it and the machine is configured for dual boot. Windows works fine on this machine and I can access the internet. When booted on Linux I get the error "Determining IP information for eth0... Failed". All these machines are connected to a Linksys 8 port router which is connected to a Linksys cable modem. I did try to power off the router and the failing computer for several minutes and then restart, but it didn't work. It worked for somebody. I know almost nothing about networking but I did look for what information others have asked and have supplied that. Hopefully someone can spot what is wrong or suggestion of where to go from here.
[root@T3304-2 ~]# cat /etc/sysconfig/network
NETWORKING=yes
HOSTNAME=T3304-2
[root@T3304-2 ~]# cat /etc/sysconfig/network-scripts/ifcfg-eth0
# nVidia Corporation MCP51 Ethernet Controller
DEVICE=eth0
BOOTPROTO=dhcp
HWADDR=00:40:CA:B1:1B:31
ONBOOT=yes
TYPE=Ethernet
[root@T3304-2 ~]# ping -c4 localhost
PING localhost.localdomain (127.0.0.1) 56(84) bytes of data.
64 bytes from localhost.localdomain (127.0.0.1): icmp_seq=1 ttl=64 time=0.212 ms
64 bytes from localhost.localdomain (127.0.0.1): icmp_seq=2 ttl=64 time=0.063 ms
64 bytes from localhost.localdomain (127.0.0.1): icmp_seq=3 ttl=64 time=0.063 ms
64 bytes from localhost.localdomain (127.0.0.1): icmp_seq=4 ttl=64 time=0.065 ms
--- localhost.localdomain ping statistics ---
4 packets transmitted, 4 received, 0% packet loss, time 2999ms
rtt min/avg/max/mdev = 0.063/0.100/0.212/0.065 ms
[root@T3304-2 ~]# ifconfig eth0 up
[root@T3304-2 ~]# ifup eth0
Determining IP information for eth0... failed.
[root@T3304-2 ~]# ifconfig -a
eth0 Link encap:Ethernet HWaddr 00:40:CA:B1:1B:31
inet6 addr: fe80::240:caff:feb1:1b31/64 Scope:Link
UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1
RX packets:31 errors:0 dropped:0 overruns:0 frame:0
TX packets:0 errors:0 dropped:0 overruns:0 carrier:0
collisions:0 txqueuelen:1000
RX bytes:5601 (5.4 KiB) TX bytes:0 (0.0 b)
Interrupt:17 Base address:0xc000
sit0 Link encap:IPv6-in-IPv4
NOARP MTU:1480 Metric:1
RX packets:0 errors:0 dropped:0 overruns:0 frame:0
TX packets:0 errors:0 dropped:0 overruns:0 carrier:0
collisions:0 txqueuelen:0
RX bytes:0 (0.0 b) TX bytes:0 (0.0 b)
**** NOTE: eth0 doesn't have a inet addr:
On Windows this is 192.168.1.102
**** NOTE: The router configuration says it will support up to 50 DHCP
I notice that your ifcfg-eth0 file specifies a hardware address. While that is permitted (I think), I don't believe it is usual, so I wonder what it is doing there. I'm wondering if for some reason the DCHP server doesn't like the specified address and therefore refuses to provide an IP address. You might try commenting out that line and see what happens.
If that doesn't work, you can monitor what happens on your network connection with tcpdump. I am guessing that when you run the ifup command there is some time elapsed before you get the error message, during which time it is trying to get a DHCP response. I would suggest first running ifdown on eth0 (I don't know whether it will give an error, but try anyway), and then in one terminal, as root, start tcpdump:
Code:
tcpdump -i eth0
Leave this running, and in another terminal, as root give the ifup command and see what information you can get. Check the tcpdumpman page for more info on this.
I commented out the hardware address line, but that didn't help.
ifdown produced no errors.
There is a long delay from ifup eth0 until the error message.
I tried the tcpdump -i eth0 and get this error message.
tcpdump: bind: Network is down
I thought that maybe I could see something from the working Fedora 5 system because from what I understand the dhcp is a broadcast request. There seems to always be some traffic even with everything shutdown I thought would access the internet. As far I can tell ifup eth0 produced no output on the working Fedora 5 system.
Fc5 stores all the stuff from ifcfg-ethx in a couple of places. Sometimes if you manually change it all the changes will not get changed everywhere. If you are running gnome I would use system->administration->network it seems to change all the stuff. You have to deactivate, make changes, hit save and then reactivate in order for it to change all the stuff in all the places. (/etc/sysconfig/network/devices/ifcg-ehx)This was a real PITA when I was bonding two cards in mode 5 (no other mode had the problem just mode 5).
I looked in /etc/sysconfig/networking/devices/ifcg-eth0 and it was an exact copy of /etc/sysconfig/network-scripts/ifcfg-eth0. The hardware address is gone now from both files. Still doesn't work. In system->administration->network there was a bootp option instead of dhcp. That didn't work either. I tried "Statically set IP addresses"
Address 192.168.1.102
Mask 255.255.255.0
Gateway 192.168.1.1
I also changed DNS screen to match a working computer. It did activate, but couldn't access the internet. It got hung in the DNS lookup. I guess because the router didn't assign the address it wasn't in the router tables. Is there some log file for networking? Some way to enable logging? Is there some way to get around dhcp like configuring manually?
I also changed DNS screen to match a working computer. It did activate, but couldn't access the internet. It got hung in the DNS lookup. I guess because the router didn't assign the address it wasn't in the router tables.
DNS doesn't per se have anything to do with the routing table. When the interface is brought up, the DNS addresses should be put in /etc/resolv.conf. Just as a result of changing the GUI, I would expect them to show up in your ifcfg-eth0 file(s) as:
DNS1=xx.xx.xx.xx
DNS2=yy.yy.yy.yy
You can find out if DNS is your only problem by pinging a website (one that is kind enough to answer pings!) using its numerical IP address. To ping www.linuxquestions.org:
ping 64.179.4.146
If the ping above doesn't work, would you post the output of ifconfig and route -n?
If you still want to investigate DHCP, if you have both Fedora boxes on the network at the same time, you can try bringing up the interface on the problematic one while running tcpdumpin promiscuous mode on the other. I believe some implementations of tcpdump automatically run in promiscuous mode and others you have to supply it as an option. Check your man page. (When it runs in promiscuous mode, you will see entries in /var/log/messages about the interface going into and coming out of promiscuous mode.
Sometimes with a dual boot when you switch between linux and windows a router reacts in a strange way.
Try resetting the router using the hidden reset button ( a reboot is not sufficient ), read the manual first, sometimes you need to press it for 10 seconds or keep the reset button pressed while you turn on the router.
The output of ifconfig and the routing table look normal to me. I don't know why you are not getting to the internet. Try arno's suggestion.
You also might try pinging between your manually configured machine and the other machines on your LAN to see if the interface really is up and the router is just getting finicky.
And just to make sure we've covered the bases, you have ruled out a firewall problem?
BTW, I noticed that unless you added that HWADDR= line back, the actual MAC address of your ethernet card matches the line that was originally in your ifcfg-etho file, so apparently that was just copied into the file at installation. That still seems to me like a strange and somewhat risky thing to do, unless I am missing something. Oh well.
If you still want to investigate DHCP, if you have both Fedora boxes on the network at the same time, you can try bringing up the interface on the problematic one while running tcpdumpin promiscuous mode on the other.
Sorry about responding to my own post, but I happened to be thinking about hubs and realized I may have left a wrong impression on my earlier post. The results you get when you put the NIC on another machine into promiscuous mode (yeah, promiscuous mode refers to the interface and not tcpdump itself ) depend upon what kind of a hub you are using. A simple hub will send everything to the monitoring machine, and you will see everything. But if you use a switching hub, you will only see the stuff that gets broadcast, which in this case I believe would be arp traffic and the original DHCP request, but not subsequent packets in the handshake (I think).
I just wanted to clear this up, both for the O.P. and anybody else who might stumble onto this thread.
Thanks for the info. I do have a switch and knew I would only see the broadcast request.
This is what I found to date. The cables are fine, I booted on Windows and it can access the internet fine. I have turned off the firewall and SELinux and it didn't help. I see the hwaddr is back. system->administration->network must have rebuilt the file(s). The hwaddr line is also on the working machine. I have reset the router several times and have even disconnected the other computers except the failing one and it didn't help. I also found from reading the router manual the IP ranges are important. If you want a fixed address it has to be in under 100. I tried this, but it didn't help either. I have also tried to ping the other machines and the ping fails, even when I use a fixed address and I can get it to activate. I'm now thinking that maybe the Linux driver uses some feature in the hardware that doesn't work. I was thinking was buying a network card. I could remove the modem card and replace it with the network card. Do you think this would do any good?
I get the same problem at thge start up and I dont have dual boot and it is for the purpose of installing Asterisk.
And I am newbie to linux and can someone tell how toinstall program and get it running as I have never ever used a linux.
I have also tried to ping the other machines and the ping fails, even when I use a fixed address and I can get it to activate. I'm now thinking that maybe the Linux driver uses some feature in the hardware that doesn't work. I was thinking was buying a network card. I could remove the modem card and replace it with the network card. Do you think this would do any good?
OK, we now have fairly clear evidence (can't ping with firewall down and with static IP address) that the interface isn't working, so the DHCP and router questions are moot for now. This is the first time I've paid attention to your hardware. You might google (and search this site) on your particular nVidia controller and see if other people are having trouble with it under Linux, and if so if there are any fixes. (Or is this the same hardware you are successfully using on your other Linux box?)
It certainly wouldn't hurt to try a network card, either as a permanent fix or as a troubleshooting technique. You would then have both eth0 and eth1, so you would have to pay attention to what you were doing. (Off hand I can't think of a way to tell Linux to pretend one interface isn't there, but I am getting tired so I might be missing something obvious. You can certainly leave one of the interfaces "down" as long as you make sure it is the right one!)
If you do try adding an NIC, I would suggest trying to get one that you know is well supported under Linux. I am familiar with 3Com's 905 series, but I know there are other well supported ones also.
EDIT: Do I understand correctly that the box you are having trouble with is the one that dual boots WinXP? If so, try unplugging the machine for a a good number of seconds (be conservative to make sure all hardware resets) and then try booting directly into FC5 and see if it works. There are reports of circumstances where MS leaves an NIC in a state where it won't work under Linux and a normal power cycle won't reset it.
@apnaveen: I cannot figure out what you are asking. Are you asking about a network problem or something else?
Last edited by blackhole54; 10-03-2006 at 06:59 AM.
LinuxQuestions.org is looking for people interested in writing
Editorials, Articles, Reviews, and more. If you'd like to contribute
content, let us know.