Help! Linux RedHat 9 won't get on Windows network
 

Ok, after reading for hours about how to fix my problem I've resorted to starting a new thread. Here's my situation.

Network:
2 PCs connected using ethernet
Both are connected to a SMC EtherEZ hub

==Machine 1==
OS: Windows ME
Comp name: irene
Workgroup: MSHOME
IP address: 169.254.0.1
Subnet Mask: 255.255.255.0
Wins config: Disabled
Gateway: none
DNS: Disabled
NetBios: none
Primary Network Logon: Client for MS Networks

==Machine 2==
OS #1: Windows 98
Comp name: CALEB
Workgroup: MSHOME
IP address: obtain Ip address automatically
Subnet Mask: " "
Wins config: Use DHCP for WINS Resolution
Gateway: none
DNS: Disabled
NetBios: none
Primary Network Logon: Client for MS Networks
OS #2: Linux RedHat 9.0

Ok, with these settings Machine 1 and Machine 2 (with OS #1) can communicate just fine. However when I tell Linux to 'Automatically obtain IP address settings with dhcp' it just hangs for a long time and says 'Determining IP information for eth0... failed.' ifconfig shows both eth0 and lo. The DNS settings have 'Hostname: localhost' and 'DNS Search Path: localhost' set and the others are blank. There are also no items in the Hosts tab.

BTW, after Linux boots up the hub light on the slot it is connected to doesn't light up. Does someone know what I can do to get them to communicate?

What kind of ethernet card do you have.

Check the output of "dmesg" to see if your card is detected during boot.

after running dmesg I found these lines that seem to be related to the network card:
PCI: Found IRQ 10 for device 00:07.0
PCI: Sharing IRQ 10 with 00:0e.0
ohci1394_0: OHCI-1394 1.1 (PCI): IRQ=[10] MMIO=[dd000000-dd0007ff] Max Packet=[2048]
ieee1394: SelfID completion called outside of bus reset!
ieee1394: Host added: Node[00:1023] GUID[00e01800000485b6] [Linux OHCI-1394]
ip_tables: (C) 2000-2002 Netfilter core team
Linux Tulip driver version 0.9.15-pre12 (Aug 9, 2002)
PCI: Found IRQ 10 for device 00:0e.0
PCI: Sharing IRQ 10 with 00:07.0
tulip0: 21041 Media table, default media 0800 (Autosense).
tulip0: 21041 media #0, 10baseT.
tulip0: 21041 media #4, 10baseT-FDX.
tulip0: 21041 media #1, 10base2.
tulip0: 21041 media #2, AUI.
divert: allocating divert_blk for eth0
eth0: Digital CD21041 Tulip rev 17 at 0xd08cd000, 21041 mode, 00:00:C0:F7:C6:E3, IRQ 10

well it seems to detect the card. What about the output of:
mii-tool -v

eth0: autonegotiation failed, link ok
product info: vendor 00:00:00, model 0 rev 0
basic mode: autonegotiation enabled
basic status: autonegotiation complete, link ok
capabilities: 10baseT-FD 10baseT-HD
advertisiong: 10baseT-FD 10baseT-HD

Considering it says that the link is OK I am surprised that the link light on the hub isn't lit. Is that still the case? Is there a link light on the network card?

The light on the card is lit (and turns off if i temporarliy remove the cable). But the hub light is still off.
Is there are way i can have it try to find any sort of connection across the network?

You could try pinging another machine on your network or the IP address of the hub if it has one - eg:
ping ip.of.some.machine

it says 'connect: network unreachable' when i type 'ping 169.254.0.1' (which is machine 1's IP address and it's light is on on the hub)

Can you post the output of:
ifconfig
route -n

here's route -n (i changed the formatting so you can read it easier):
Destination = 127.0.0.0
Gateway = 0.0.0.0
Genmask = 255.0.0.0
Flags = U
Metric = 0
Ref = 0
Use = 0
Iface = lo

here's ifconfig:
eth0
Link encap:Ethernet HWaddr 00:00:C0:F7:C6:E3
UP BREADCAST RUNNING MULTICAST MTU:1500 Metric:1
RX packets:0 errors:0 dropped:0 overruns:0 carrier:34
collisioins:0 txqueuelen:100
RX bytes:0 (0.0 b) TX bytes:0 (0.0 b)
Interrupt:10 Base address:0xd000

lo
Link encap:Local Loopback
inet addr:127.0.0.1 Mask:255.0.0.0
UP LOOPBACK RUNNING MTU:16436 Metric:1
RX packets:654310 errors:0 dropped:0 overruns:0 frame:0
TX packets:65431 errors:0 dropped:0 overuns:0 carrier:0
collisions:0 txqueuelen:0
RX bytes:4535663 (4.3 Mb) TX bytes:4535663 (4.3 Mb)

Is that all that was in "route -n"? if you post it in [CODE] blocks the formatting isn't lost.

Here's the 'route -n' output again:

Kernel IP routing table
Destination Gateway Genmask Flags Metric Ref Use Iface
127.0.0.0 0.0.0.0 255.0.0.0 U 0 0 0 lo
You have new mail in /var/spool/mail/root

disable the auto configuration of ur ethernet and give the manual setting and assign ip address in windows wht ever ur using dont not configure any dns server use the command setup after logged in as a root to configure your ethernet card

hope.....helpful

Do what satish says. Here's why.

Your first 'puter has a "bogus" APIPA subnet mask: it should be 255.255.0.0. When you tell your second 'puter in Win98 to obtain an IP address automatically, that's the mask it will get. So your first and second 'puters are not on the same subnet.

As for Linux, I don't think it does APIPA. If there is no DHCP server, it is not going to get an IP address. So you'll have to assign it a static IP address.

I've tried going into the Network configuration and staticly setting 'Address=194.254.0.2' 'Subnet mask=(blank)' and 'Default Gateway Address=255.255.0.0' in Linux. Now when I do a 'route -n' it says:
Destination Gateway Genmask Flags Metric Ref Use Iface
194.254.0.0 0.0.0.0 255.255.0.0 U 0 0 0 eth0
169.254.0.0 0.0.0.0 255.255.0.0 U 0 0 0 eth0
127.0.0.0 0.0.0.0 255.0.0.0 U 0 0 0 lo

And when I try 'ping 169.254.0.1' it outputs:
PING 169.254.0.1 (169.254.0.1) 56(84) bytes of data.
From 169.254.0.2 icmp_seq=1 Destination Host Unreachable
From 169.254.0.2 icmp_seq=2 Destination Host Unreachable
From 169.254.0.2 icmp_seq=3 Destination Host Unreachable
...........etc.

Sorry, but i'm not incredibly network savy. So I'm not sure what exactly needs to be disabled to remove the autoconfiguration on the network.

BTW the light on the hub where the cable from the linux box is going to still isn't on.
Thanks for anyones help.

I've just noticed you are using very odd IP addressesand you need a mask - try:
Machine 1:
IP: 192.168.1.5
MASK: 255.255.255.0

Machine 2:
IP: 192.168.1.6
MASK: 255.255.255.0

If they aren't conencted to the internet there is no need to specify a gateway. If they are then use the IP address of the computer with the internet connection.

BTW when i boot up linux it gives me a message window saying:
"Could not look up internet address for localhost. This will prevent GNOME from operating correctly. It may be possible to correct the problem by adding localhost to the file /etc/hosts"

My /etc/hosts file only has two commented out lines in it and that is it.

My 'DNS' and 'Hosts' tabs don't have anything in thier fields in RedHat.

I've set the IP and Masks to what you suggested and both machines have no problem talking to eachother in windows but linux still does even with those settings.

Is that hub light still off? I guess normally you'd think the hub and card would either both be on or off, but if you mistakenly had a crossover cable and the card autodetected... I think I'd try another cable.

The thing is that cable works just fine when I boot with Windows 98 rather than Linux. I'm beggining to wonder if Linux didn't detect the network card correctly. But I don't get a [FAILED] response in the Linux boot sequence when it loads eth0. Does the fact that the /etc/hosts file is empty have anything to do with it?

/etc/hosts should have at least:
127.0.0.1 localhost localhost.localdomain

Temujin_12,

FYI - As david_ross said, you can put the stuff you copy from your terminal within the [ code ] tags to preserve the formatting.

For instance:
[ code ]
whatever you copy and paste goes here
[ /code ]
But, don't put the spaces after the '[' or before the ']'. I did that here just so you could see how they are used.

Example:

I've forgotten what is important, but my /etc/hosts has:

127.0.0.1 localhost.localdomain localhost
192.168.1.nn myname myname.myworkgroup (except with real numbers and names)

The rest of the office is on DHCP, but I'm stuck with a fixed IP because the terminal emulator they use to get to me requires it.

*arg* I just read this thread for the first time and I notice quite a few small problems... I wish I had seen it earlier. Lets start over and combine all the suggestions above.

With regards to your original configuration, you have your machines setup with DHCP but you don't have a DHCP server! *ah* no wonder you don't get an IP back. Microsoft takes the liberty of assigning a 169.254.0.x address when no DHCP server is found where Linux, more appropriately, fails to assign an ip address when no DHCP server is found - this is the cause of your "fail" when starting that adapter.

Secondly, when you setup your computers such that machine 1 is on the 169 network, machine 2 is on the 194 network... they won't talk.

So... before you check anything else, lets get your network configured correctly. Either install a DHCP server into your network -OR- (and a better option for testing) setup your machines with static ip addresses. i.e. Machine 1, 192.168.0.2 & Machine 192.168.0.3 - Notice that they are both on the 192.168.0.0/24 network.

Once that is done... go back to the default routing table. If you have trouble here, let us know and we'll help you sort it out.

Lastly, make sure your firewall isn't being overly restrictive and blocking all communication.

Keep us posted and we'll continue from there!
Cheers,
J.

I don't doubt you a minute, but I still don't understand. ifconfig was used to set a fixed address to match the Windows DHCP defaults, yet it didn't work. And what did any of that have to do with the hub light being off?

BTW, A cheap router between your modem and LAN will demand to assign the addresses, if that is in the forseeable future.

gernt, you are right that the hub light remaining off sounds like a problem. Have you tried plugging the cable into a different port on the hub?

Everything else... well... All those symptoms are indicitive of a network configuration problem. I'm not convinced that his linux box has been configured with the correct info to be on the same network as his Windows machine.

Here's some things we need to look at, please re-post your current...
1. Your ifcfg-eth0 file
2. Your /etc/sysconfig/network file (should include NETWORKING=yes)
3. Results from "ping 127.0.0.1" on your linux box
4. Results from "route -n" on your linux box
5. Results from "iptables -L"

Thanks,
J.

Ok, here is what I've done.

1) Changed my TCP/IP IP to '192.168.0.2' with a subnet mask of '255.255.255.0' on machine 1.
2) In linux I went to System Tools >> Network Device Control and set it to 'Statically set IP addresses' and set them to: IP address '192.168.0.3' Subnet Mask '255.255.255.0'

Still no light on the hub....
BTW...
And here are my outputs from the commands and files JordanH asked me to post (I'll try to use the [ code ] command)

It doesn't sound to me like this network has a DHCP server. Setting one up would be the optimum solution, I would think.

I know this must be exasperating, but if you get ping to work, the rest follows. One thing you can try is to get a crossover cable and connect two computers directly without the hub. (Label the crossover cable clearly - otherwise it isn't obvious and you end up putting it where it shouldn't go). Please be sure the IP address is still OK by doing an ifconfig. It should show eth0 at the expected address. Then you can try to ping 192.168.0.2 from the linux machine, and 192.168.0.3 from the other. If this still fails, maybe your card is not compatible with linux.

Your configs look good now.

When you restart your network, does it still fail to bringup the eth0 card?

I know this is an old post but just in case anyone reads through it with a similar problem I wanted to follow up and write down what my problem was.

It turned out that the way my NIC was being called via RedHat caused it to be unable to communicate with my SMC EtherEZ Hub 3605T (Ethernet Hub). When I hooked it up to another router (or directly into my modem) everything worked fine.

The strange part was that Windows would communicate with it just fine. I suspect that it was some sort of low level service call error.