Networking is not working

I cannot get my newly installed red hat 8 system to talk on the internet. It appears to be giving arp's (ethereal), but thats it. It can ping 127.0.0.1 and 192.168.2.2 (it's ip address) but it can't ping 192.168.2.1 (the default gateway). I have the subnet mask setup correctly, and I know this should work because I was just using my (used to be) windows box with the same cable, ip address, etc. So it appears to me that the card is functioning to some extent. The card starts up fine on boot, and I can activate and deactivate it once logged in. Anyone that can help me would be much appreciated. Thanks,

Jon

Are there any firewall setup that might cause the problem? Try to run 'iptables -v -L' to see if there are some blocking.

Hi -

You're asking all the right questions:
1. You believe your IP address and subnet mask are configured correctly.
2. You can ping "127.0.0.1" (so the TCP stack itself is working)
3. You *can't* ping your gateway (so of course you won't be able to talk to anybody beyond that).

So the two questions - the *only* two questions at this point - are:
1. Is Linux talking to your NIC (if not, that's the problem)
... and, if so, then ...
2. Why can't you "ping" the gateway

Please try this:

1. run "netstat -in"
EXAMPLE:
Kernel Interface table
Iface MTU Met RX-OK RX-ERR RX-DRP RX-OVR TX-OK TX-ERR TX-DRP TX-OVR Flg
eth0 1500 0 52 0 0 0 55 0 0 0 BMRU
lo 16436 0 126 0 0 0 126 0 0 0 LRU
<= SO THE STACK IS SEEING INTERFACE "ETH0"...

2. Run "ifconfig IF_NAME"
EXAMPLE:
/sbin/ifconfig eth0
eth0 Link encap:Ethernet HWaddr 00:03:6D:10:37:A3
inet addr:192.168.100.2 Bcast:192.168.100.255 Mask:255.255.255.0
inet6 addr: fe80::203:6dff:fe10:37a3/64 Scope:Link
UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1
RX packets:52 errors:0 dropped:0 overruns:0 frame:0
TX packets:55 errors:0 dropped:0 overruns:0 carrier:0
collisions:0
RX bytes:7185 (7.0 Kb) TX bytes:6413 (6.2 Kb)
<= THIS VERIFIES WE'RE TALKING TO THE NIC (I.E. WE SEE THE CORRECT
MAC ADDRESS), AND THE SUBNET AND IP ARE THE ONES WE EXPECT

3. If everything checks out so far, look at your startup errors (e.g. in /var/log/messages)

4. If you *still* can't find the root of the problem, consider deleting the interface from your configuration, rebooting, and re-adding it again.

Your .. PSM

I checked the iptables and nothing seems to be blocking. Earlier I shutdown iptables to make sure that wasn't a problem. I ran netstat -in and it found my eth0. Looking in my error log, everytime I tried to transmit, it negotiated a half-duplex, but that should be fine, since my card is a 10/100 and I have a 10 mbps hub its hooked into. I started with one card (netgear FA311) and that didn't work (same symptoms) so I switched to the same make/model nic that is in my redhat 8 server, so I figured it would work fine. I will delete the hardware and restart and post back.

Thanks for your help.

Jon

I went ahead and deleted it from the configuration, rebooted, readded it. I look at the settings from my server with the same nic and they look identical.

The only difference between the two is the machine and the load. I did a workstation load on the one that isn't working (and added a few packages), and a server load on the one that is working. My server is on a AMD K6-2 500 mhz (I forget what kind of board) and my workstation is a soyo celeron 2.0 with the latest bios updates.

What is the model of your nic? What kernel are you using? (I have had some strange problems with via-rhine and broadcom cards with earier kernels)

Does the leds behind your nic blink when you try to ping your gateway? Ping just doesn't return or does it give some error message (like "network unreachable" or similar)?

Is this gateway an other computer or some box to talk with? Do you have a chance to verify that the gateway is answering to pings?
Do you have cross-connected cable which you could use to temporarily connect this machine to the gateway to eliminate the possibility that there is some problem with talking to the hub? You could also try to use 'mii-tool' to force your nic to certain modes.

I am using kernel-2.4.18-14 The NIC is either (I can't remember which one I put in) Gigafast Ethernet Adapter model EE100-AXP or a Belkin model P56632-B If it's the Belkin, (I think that is the model number). If it is the belkin, its the exact same card as in my server. They are both using the same driver, so I think it is the same one. They are both cheap cards with not much to them. I also had a netgear FA311 in there initially. I know it isn't the hub/cables because this was a windows box that the networking worked fine, then it turned to a linux box and this isn't working.

When I ping the NIC doesn't blink and I get Destination Host Unreachable. Looks like info isn't leaving the nic. I hooked up my server to this subnet and ran ethereal and it only picked up arp requests from the client. The gateway is a computer setup with smoothwall.

I don't think I've ever used the mii-tool, how do I go about that?

Thanks,

Jon

EXAMPLE:
tiny:~ # mii-tool -v eth0
eth0: no autonegotiation, 10baseT-HD, link ok
product info: vendor 00:08:95, model 1 rev 0
basic mode: autonegotiation enabled
basic status: autonegotiation complete, link ok
capabilities: 100baseTx-FD 100baseTx-HD 10baseT-FD 10baseT-HD
advertising: 100baseTx-FD 100baseTx-HD 10baseT-FD 10baseT-HD flow-control
link partner: 10baseT-HD

You can do a "man mii-tool" for more details.

Did you say you have one RH8 distro (server) that's working, and another RH8 (workstation) that isn't? Have you tried doing an ethereal capture of a "ping" from the working host to the failing host?

Another suggestion - please try "ifconfig eth0 up" (substituting your interface for "eth0")
Who knows - it couldn't hurt...

Ok, I forced it to 100 full and half, and 10 full and half and no luck. I did an ethereal capture and all it saw was arp's from the client that isn't working. So it appears there is some communication going out, but just the arp's. There weren't any ARP replies. I think I'm gonna wait until tomorrow and try to get redhat 9 to upgrade to and or compile a new kernel (will be my first attempt at doing so).

Thanks for your help, I'll let you know of the outcome.

Thanks,
Jon

Ok, I switched out the card I know is working from my server and put it in the client that wasn't working. I still got the same results, the new card I put in the server started working, and the card I put in the client isn't working. I then installed fedora hoping that would help, but it didn't. I tried to boot knoppix to try there, and that didn't work either. I then tried to connect with my modem and that didn't work, so I am suspecting something with my whole pci bus (those are the only two devices there). I have a soyo p4-is2 motherboard with a celeron 2.0. I am wondering if the chipset isn't completely compatible with the kernel? But the devices get recognized so I know it is working somewhat, and I can ping 127.0.0.1 and 192.168.2.2 (its own ip address). I have tried upgrading the kernel but was unsuccessful, since I have never done it before. I will try applying the patches or keep trying to upgrade the kernel to see if that helps. I am out of ideas, as I am fairly new to linux. Any help would be appreciated.

Thanks,

Jon

sorry to but in, but i have a similar problem.
i have Java Desktop installed, my eth0 is sending packets,but not receiving any replies.

i have a log error:
linux kernel: eth0: no IPv6 routers present

when i run ifconfig eth0
if get
Link encap:Ethernet HWaddr 00:09:5B:60:47:8E
inet addr: 192.168.100.64 Bcast:192.168.100.255 Mask: 255.255.255.0
inet6 addr: fe80::209:5bff:fe60:478e/64 Scope:Link
IP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1
RX packets:0 errors:0 dropped:283 overruns:0 frame:0
TX packets:0 errors:0 dropped:0 overruns:0 frame:0
collisions: 0 txqueuelen: 100
RX bytes :0 (0.0 b) TX bytes: 0 (0.0 b)
Interrupt:11 Base address: 0x5000

I also have a red hat 7.0 running on this network, but it makes no reference to IPv6 when i run ifconfig eth0
I have a new router (linksys) that runs everything else fine. The cables, hubs and ethernet card on this computer all check out fine and each runs on other computers ok.

any ideas how i can make this work?

thanks stephen

Hi, Jon -

Sorry - I honestly thought you were already squared away.

It *should* work (you say it used to work on Windows; "netstat" and "ifconfig" show that it seems to have detected the H/W and found an appropriate kernel driver, etc etc), or, at this point, it should fail with a conspicuous, meaningful error (i.e. a device error at boot time in /var/log/messages).

I'd suggest the following:
1. Create a new post (i.e. put yourself at the top of the list)

2. Cable directly from your failing PC into one of your working PCs
(no routers, no hubs; short, known-good cables, etc). Focus on
getting connectivity *just between those two hosts*.

3. Briefly describe your two PCs: their h/w, the OS and OS versions,
the NICs, exactly what works, exactly what doesn't work ... and "what
changed".

4. Try to focus on one variable at a time.

I look forward to seeing your new post, and seeing what "additional eyes" might discover.

Please check out these FAQs, too:

- http://www.linuxquestions.org/questi...hreadid=173280

- http://www.linuxgazette.com/node/view/4186

- http://jeays.net/files/rh9-linksys.htm

This may sound stupid, but check the cable, and watch the LED light on the nic, and switch, when you do a ping, and see if they blink.
also check your routing table. none of the post here seem to mention anything about routing table. so, I just have to thrown it it there. Unlike, windows, once you assign the ip, it doesn't neccessary add the local network 192.168.2.0/24 to your routing table.

Originally posted by jonlake Ok, I switched out the card I know is working from my server and put it in the client that wasn't working. I still got the same results, the new card I put in the server started working, and the card I put in the client isn't working. I then installed fedora hoping that would help, but it didn't. I tried to boot knoppix to try there, and that didn't work either. I then tried to connect with my modem and that didn't work, so I am suspecting something with my whole pci bus (those are the only two devices there). I have a soyo p4-is2 motherboard with a celeron 2.0. I am wondering if the chipset isn't completely compatible with the kernel? But the devices get recognized so I know it is working somewhat, and I can ping 127.0.0.1 and 192.168.2.2 (its own ip address). I have tried upgrading the kernel but was unsuccessful, since I have never done it before. I will try applying the patches or keep trying to upgrade the kernel to see if that helps. I am out of ideas, as I am fairly new to linux. Any help would be appreciated. Thanks, Jon

Hi -

Actually, your suggestion isn't at all stupid. It's very, very intelligent. Jon absolutely needs to "double-check the obvious". Doube-check the cabling, double-check whether the NIC light blinks, double-check the H/W compatibility lists (HCL) ... take an Ethereal (Linux) or Network Monitor (Windows) package trace from *both* sides of the connection ... perhaps even consider just buying a new NIC ...