Help in setting up a red hat gateway

Cirvam · Nov 23, 1999

Hi 
 
I have a 486/DX (66mhz) with 40 mges of ram and a full install of Red Hat 5.1. I have a NE2000 compatible network card in and have that configured (I think). I have two other Win95 machines on the same network. The two Win95 machines can ping each other but the Red hat machine can't communate with any thing except its own loopback address and the IP I assigned the machine. The other machines all have IPs and I have a route to the network but it won't work. 
 
thanks for you help 
Erik

sijones · Nov 24, 1999

make sure you put in the hosts file your other win95 ip - names, this will make ping work. file can be found in 
 
/etc/hosts 
 
just edit this and should be OK. if you type route you should get an output with at least one address but should be more, you should have the lo interface as 127.0.0.1 then the address of the machine it self eg. 192.168.0.3 this should go with your eth0 interface. 
 
Hope this helps! Simon Jones <a href=mailto:sijones@talk21.com>sijones@talk21.com</a> <a href=

http://www.theallandale.co.uk>The

Allandale Youth Works</a> The above is all my own site I look after!!! 
God am good!!! and good lookin

Cirvam · Nov 24, 1999

I have everything added to the hosts file and my route looks like this: 
Destination Gateway Genmask Flags Metric Ref Use Iface 
192.168.1.0 * 255. 255.255.0 U 0 0 0 eth0 
127.0.0.0 * 255.0.0.0 U 0 0 1 lo 
 
And it still doesn't work

Eventually I want to have every computer use this as the dial out hub but right now I just want networking to work. If I ping the IP I assigned to it (192.168.1.1) then the pings come back. If I try to ping any of the other computers it just sits there, it doesn't even say host unreachable. If I use a win95 machine to ping the Linux box all requests time out. 
 
Thanks 
Erik

sijones · Nov 25, 1999

I have looked at my route and it has four routes, 
 
ip gateway netmask 
192.168.0.3 * 255.255.255.255 eth0 
192.168.0.0 * 255.255.255.0 eth0 
local one here 
default allandale 0.0.0.0 eth0 
 
the allandale name on this route is for internet access, I have put allandale in my hosts file, and it just works for me and can ping a win98 and nt4 boxes from my linux box. 
 
Carn't think of anything else that might stop it but I will look up in my rh6 book on it and see what it might be. 
 
Try adding the routes as above but with your ip addresses, 
look at the route command help to see how to add a route. Simon Jones <a href=mailto:sijones_uk@yahoo.com>sijones_uk@yahoo.com</a> <a href=

http://www.theallandale.co.uk>The

Allandale Youth Works</a> God am good!!! and good lookin

AndyBo · Nov 26, 1999

First tip - ping on RedHat (I think 6.0, but may be 5.2 onwards) is broken. Retrograde your ping with an earlier version such as <A HREF="

http://www.rpmfind.net/linux/RPM/redhat/6.0/i386/netkit-base-0.10-29.i386.html"

TARGET="_new">

http://www.rpmfind.net/linux/RPM/redhat/6.0/i386/netkit-base-0.10-29.i386.html</A>

(Problems in the newer versions include not timing out when a host can't be contacted. What are they thinking of?) 
 
What are the IP addresses of -all- your boxes? What I'm thinking is are they all on the 192.168.1.0 network? If they are, you don't need to go around adding routes, etc. Devices on the same network should all be able to contact all other devices without adding to routing tables. In fact, if you take a look at your routing tables before you start adding anything, you should see something like: 
 
ip gateway netmask 
192.168.1.0 * 255.255.255.0 eth0 
 
(As an aside, the route in sijones' routing table to 192.168.0.3 is a route to a host - not a network. ie, the routing table is saying "To get to this specific host, use interface eth0, and a netmask of 255.255.255.255". How this works, I don't know, cos the 255.255.255.255 netmask is saying that the entire IP address is a network address, not a device address. Hmmm....) 
 
On the physical side of things, what sort of cabling are you using? Twisted pair (RJ 45), AUI, or co-ax? If it's co-ax, try using a different cable segment for the Linux box. If it's TP, try switching the ports around in your hub. ie, plug the Linux box into a port currently used by a Windows box. 
 
Hope some of this helps

Cirvam · Nov 26, 1999

I have 3 hosts on the network: 
192.168.1.1 - the Red Hat box 
192.168.1.2 - one Win95 
192.168.1.3 - the other Win95 
 
I just re-installed Red Hat and checked all the cables(Twisted Pair) and I switched ports on the hub and I switched the network card in the Linux box. I think the ping problem is there but the ping still works if it times out, also I can't ping from a Win95 machine to the Linux one. 
 
Right now I am downloading Red Hat 6.1, I'm going to try and use that and see if that changes anything. 
 
Thanks 
Erik

warmongr · Nov 27, 1999

From my desk, I can't tell you what's wrong, but I know it's not RH6.0 as I have multiple gateways established using 3c90x cards from 3Com and Rh6. Do you have the correct driver or module for your card? However to get to two different physical networks we use two network cards in our gateways configuring one as eth0 and the other as eth1. As AndyBo says the routes don't need to change on the same network and in order for multicast to work the default routes should be themselves. 
 
do a dmesg¦more and check for the loading of your nic driver. 
 
war// 
 
AndyBo: Do you have any supporting references for your ping problem. I would like to check this out for future reference. 
 
Thanks. 
war

sijones · Nov 29, 1999

Ping works on rh6, that I use. Simon Jones <a href=mailto:sijones_uk@yahoo.com>sijones_uk@yahoo.com</a> <a href=

http://www.theallandale.co.uk>The

Allandale Youth Works</a> God am good!!! and good lookin

AndyBo · Nov 29, 1999

Re the ping problem - this has come up a few times on the Big Brother mailing list. (Big Brother is a most excellent system monitoring tool available from <A HREF="

http://www.bb4.com"

TARGET="_new">

http://www.bb4.com</A>

If you need to monitor more than a couple of boxes and services, it comes highly recommended.) 
 
Digging through my mail archives... A few quotes... 
 
"Since this is a RedHat 6.1 system, you should definitely try reverting 
ping to an earlier version, there's a known (and reported) bug in the new 
ping from RH6.1 (and updates to 6.0) where a missed response will cause 
the silly thing to hang almost infinitely instead of stopping after a 
timeout." 
 
And in a little more detail from the same correspondent later... 
 
"I'll reply to this one again, since I didn't give any details. 
 
The bug as originally reported only shows when you ping a host not on 
your local ip net, ie. one where you ping through a router. 
 
It's still present in the newest rpm, and when I check Bugzilla, the 
bugreport for the problem is still listed as NEW, not as VERIFIED or 
FIXED, so the RH haven't even looked at it yet. 
 
Just to show the behaviour: 
root@cyril:/tmp# ./ping -c 1 10.10.10.10 
PING 10.10.10.10 (10.10.10.10) from 10.0.18.64 : 56(84) bytes of data. 
64 bytes from 10.10.10.10: icmp_seq=0 ttl=255 time=0.8 ms 
 
--- 10.10.10.10 ping statistics --- 
1 packets transmitted, 1 packets received, 0% packet loss 
round-trip min/avg/max = 0.8/0.8/0.8 ms 
This worked. 
 
root@cyril:/tmp# ./ping -c 1 10.0.0.2 
PING 10.0.0.2 (10.0.0.2) from 10.0.18.64 : 56(84) bytes of data. 
From cyril.iaeste.dk (10.0.18.64): Destination Host Unreachable 
 
--- 10.0.0.2 ping statistics --- 
1 packets transmitted, 0 packets received, +1 errors, 100% packet loss 
This also worked, though the time before giving up was ~3 seconds instead 
of the ~10 in the old version. 
 
root@cyril:/tmp# ./ping -c 1 130.245.25.5 
PING 130.245.25.5 (130.245.25.5) from 10.0.18.64 : 56(84) bytes of data. 
^C pressed 
 
--- 130.245.25.5 ping statistics --- 
4 packets transmitted, 0 packets received, 100% packet loss 
 
This breaks badly, note that it tried multiple packets, and never timed 
out, so is utterly useless for us." 
 
In other words, when trying to ping a non-existent host on the local network, you're told about it. When trying to ping a host on a different network, you aren't. Note the "-c 1" to send 1 packet in all examples, yet in the last example it just keeps pinging away. 
 
Doesn't sound like this is the problem with Cirvam, though. I think War has hit the nail on the head with NIC modules not being loaded, though. 
 
Hope this helps.

Cirvam · Nov 29, 1999

I looked back at the boot records and it said the drivers were loading. I recompiled the kernal and have to reconfige the interuppets. But before the modules were loading and the only strange thing was sendmail would hang when booting, but everything else looked normal. Any other suggestions?

Erik

AndyBo · Nov 30, 1999

Once the system's been up a few minutes, what does 'lsmod' show? Is the NIC module still loaded? Also, is the module loading with the right IO address, IRQ, etc? Is it a PnP NIC? That may cause a problem. Do you have any DOS based configuration programs that you can use to interrogate the card with? (ie, to check what it's settings are.) 
 
Is the card in the server the same model as the cards in the Win boxes? If not, could you swap the cards to see if that makes a difference? 
 
Let us know

Tek-Tips is the largest IT community on the Internet today!

Members share and learn making Tek-Tips Forums the best source of peer-reviewed technical information on the Internet!

Help in setting up a red hat gateway

Cirvam

Programmer

sijones

Programmer

Cirvam

Programmer

sijones

Programmer

AndyBo

MIS

Cirvam

Programmer

warmongr

MIS

sijones

Programmer

AndyBo

MIS

Cirvam

Programmer

AndyBo

MIS

Similar threads

Part and Inventory Search

Sponsor