I decided to see if I could figure out how to set up a very basic home network and have run into some problems which I can find no way around. I am using two older computers which I had lying around and have never had any use for. I put fresh installs of Slackware on them, and so there is no firewalls or any such thing on them. I put in a couple of similarly old ISA ethernet cards (first SMC EtherEZ and then 3Com Etherlink IIIs) which I had salvaged from various sources. I figured that I would work with what I had to see if I could get something going before using newer cards on my regularly used computers.

I ran modprobe and the cards were apparently recognized and the drivers seemed to load fine in all cases. I used only one card per machine at a time by the way, but switched them out after the first ones failed. I ran netconfig and set up the IP addresses as was suggested by an online site. When I ran ifconfig on the first computer to check things, this is what I found:

eth0 Link encap:Ethernet HWaddr 00:00:C0:06:B8:EB
inet addr:192.168.1.1 Bcast:192.168.1.255 Mask:255.255.255.0
UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1
RX packets:5 errors:0 dropped:0 overruns:0 frame:0
TX packets:169 errors:0 dropped:0 overruns:0 carrier:0
collisions:0 txqueuelen:100
RX bytes:300 (300.0 b) TX bytes:10482 (10.2 Kb)
Interrupt:3 Base address:0x250 Memory:c0000-c2000

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:147 errors:0 dropped:0 overruns:0 frame:0
TX packets:147 errors:0 dropped:0 overruns:0 carrier:0
collisions:0 txqueuelen:0
RX bytes:17184 (16.7 Kb) TX bytes:17184 (16.7 Kb)

From what I have found online, this appeared to be okay. The results for 192.168.1.2 were basically the same.

I am only using the two machines to start with and so used a crossover cable to connect them. I expected things to go pretty smoothly at this point, but I could not actually make any connection. 192.168.1.1 could ping itself, but not 192.168.1.2, with the same being true in reverse. When I tried to ping the other machine this is what I got:

PING 192.168.1.2 (192.168.1.2) 56(84) bytes of data.
From 192.168.1.1 icmp_seq=1 Destination Host Unreachable
From 192.168.1.1 icmp_seq=2 Destination Host Unreachable
From 192.168.1.1 icmp_seq=4 Destination Host Unreachable
From 192.168.1.1 icmp_seq=5 Destination Host Unreachable
From 192.168.1.1 icmp_seq=7 Destination Host Unreachable
From 192.168.1.1 icmp_seq=8 Destination Host Unreachable
From 192.168.1.1 icmp_seq=10 Destination Host Unreachable
From 192.168.1.1 icmp_seq=11 Destination Host Unreachable

--- 192.168.1.2 ping statistics ---
13 packets transmitted, 0 received, +8 errors, 100% packet loss, time 12039ms
, pipe 2


The same results were found from the other computer, and with the other cards as well.

I am wondering what I am either overlooking or doing wrong? The online guides I have found mentioned no other scripts or such which needed to be configured to have a successful result and so I am a bit lost. I am definitely using a crossover cable, and so that should be okay. Could the old cards be to blame? I thought that may be the case at first, which is why I switched from the SMC to the 3Com cards, but found no change at all.

Any help at all or suggestions is greatly appreciated. I would just like to know if I can make something work before spending any money, being that I am such a cheapskate. :-)

What does
route
look like on both of them?

You should make one of them the default gateweay (even
if you don't have an outside interface).



Cheers,
Tink

Okay, I went ahead and ran netconfig again and put in the first computer as the gateway. I then ran route and got this:

Kernel IP routing table
Destination Gateway Genmask Flags Metric Ref Use Iface
localnet * 255.255.255.0 U 0 0 0 eth0
loopback * 255.0.0.0 U 0 0 0 lo
default 192.168.1.1 0.0.0.0 UG 1 0 0 eth0

And just for a lark I tried to ping again and got the same as before. What do you think?

Hard to say what else might be borked ...

Maybe one of the cards is trying to talk
10, the other 100MBit, duplex vs none, ...

Do you know for sure that the cable is
good?


Cheers,
Tink

How could I check those things? I have no way of testing the cable since I don't have any other networking setup going, this being my first. And I have no knowledge of the cards or how to adjust things on them? Are the things you mentioned, i.e. duplex etc, adjustable or checkable?

Many thanks again.

Some of these settings can be modified by module.config
if the drivers are compiled as modules (and of course it
depends on the actual driver in use - not all drivers are
designed in an indentical manner). Have a look at the
relevant files in your kernels-source tree ... the
<path/to/kernel-source>/Documentation directory provides
lots of info on stuff like this.

As far as the testing goes - means are quite limited.
You could try to find someone who has a working
set-up to test your cable.


Cheers,
Tink

>You should make one of them the default gateweay
>(even if you don't have an outside interface).

No, TCP/IP in linux works without a default gateway,
you don't need a default gateway unless you have connection
to another subnet or the internet.

>What do you think?

What version of slackware is that, the output of route on 9.1 looks slightly different.

It sound like you have done everything right so far. I suspect a hardware problem or a problem with the driver modules.

Do the leds on the networks cards flash when they should be sending packets.

Run tcpdump -i eth0 on both te machine and see if either of them gets packets from the other when you ping.

Well, I have had some success, I think. In doing some very tedious googling I came across some people who had the SMC EtherEZ cards and one mentioned the irq and io he used. I passed these, along with some info which I found in other sites discussing different cards, to lilo at boot in hopes it would help, and amazingly it seemed to actually do that. Specifically, I used the string "ether=9,0x240,4,0,eth0" and now I do get a successful ping, or at least it appears so, from each computer to the other.

This apparent success leaves me wondering about some things though. For instance, how do I use this autmatically? Can I use an append in lilo.conf?

And what do I do next to actually use the connection? What is required to actually access the hard drive, for instance, on one machine from the other? The stuff I am finding online seems to take for granted some idea of what one is looking to do, whereas admittedly I am really in the dark about it altogether.

And many thanks for all the help.

Append in lilo would work, but the prefered method
would be to add it as an option in /etc/modules.conf


Cheers,
Tink

Many thanks. I will look into the way to go about doing that.