I currently have a redhat 6.2 linux box sitting on a dmz on our lan. It has always been the only box on the dmz. The firewall it uses as its gateway is 192.168.1.1. It's IP is 192.168.1.2. They are both static IP addresses. I use the box for our primary external DNS as well as mailforwarding using sendmail. It has always worked perfectly but it's time for an upgrade.

I got the new box, installed Mandrake 9.2, postfix, amavisd, bind, etc. and set it up to completely mimic the existing server. I pulled the old machine off the network and shut it down, put the new server on the network using the same IP, subnet, gateway, etc of the old server, and even used the same network cable, and I can't ping ANYTHING. Even itself using 127.0.0.1 or its IP of 192.168.1.2.

To test, I took the server off the DMZ and put it on our 10.1.x.x network and gave it a static IP and I'm able to ping anything I want.

I also, just for kicks, put the old machine back online at its original 192.168.1.2 address and changed the new machine to 192.168.1.3 and I was able to ping the old machine, but still not itself, or the gateway.

Any ideas? It could be something simple, I'm fairly new to linux. LOL

Anything is appreciated.

Thanks in advance!

Welcome to LQ.

Can you psot the output of:
route -n
ifconfig

The route -n looks like this... sorry I couldn't copy it..
IP gateway mask interface
192.168.1.0 0.0.0.0 255.255.255.0 eth0
10.1.26.0 0.0.0.0 255.255.255.0 eth0
127.0.0.0 0.0.0.0 255.0.0.0 lo



The ifconfig looks like this.... (just the important stuff..)

eth0 Link encap: Ethernet
Inet addr: 192.168.1.3 Bcast: 10.1.26.255 Mask: 255.255.255.0
UP BROADCAST RUNNING MULTICAST
RX Packets:5460 errors:0 dropped:0
TX Packets:148 errors:0 dropped:0


lo Link encap: Local Loopback
inet addr: 127.0.0.1 Mask:255.0.0.0
UP LOOPBACK RUNNING


Looks like I've got some problems. But I don't know how to change them. :-)

Take a look at your eth0 config file and set the correct broadcast and gateway:
/etc/sysconfig/network-scripts/ifcfg-eth0

I fixed the items in the ifcfg-eth0 file. I rebooted afterwards and it shows up correctly when I look at ifconfig, but I'm still having exactly the same problem. I am totally stumped.

Thanks for the help, and keep the ideas coming.

I REALLY appreciate the help.

What does the output of the commands I posted before show now?

The route -n looks like this...IP gateway mask interface

192.168.1.0_____ 0.0.0.0_____ 255.255.255.0_____ eth0
127.0.0.0_______ 0.0.0.0_____ 255.0.0.0_________ lo
0.0.0.0_________ 192.168.1.1__ 255.255.255.0_____ eth0


The ifconfig looks like this.... (just the important stuff..)

eth0 Link encap: Ethernet
Inet addr: 192.168.1.3 Bcast: 192.168.1.255 Mask: 255.255.255.0
UP BROADCAST RUNNING MULTICAST
RX Packets:5460 errors:0 dropped:0
TX Packets:148 errors:0 dropped:0


lo Link encap: Local Loopback
inet addr: 127.0.0.1 Mask:255.0.0.0
UP LOOPBACK RUNNING

Well you have a default route now but the mask is way of - try deleting it:
route del default

Then adding the proper one:
route add default netmask 0.0.0.0 gw 192.168.1.1 dev eth0

still didn't work.. giving up for the night.

Thanks for all the help!

Do you have a firewall turned on? It is possible to set a firewall to where you cannot ping anything, including 127.0.0.1. If you type iptables -L it will give you your firewall definitions. If you do have one on, the turn it off and try to ping.

Ok, thanks. I'll look at that, but I did put the box on the 10.x.x.x network and it worked fine, so I wouldn't think that would be it.

I'll definately look at it though.

Another idea.... I'm from the windows world, so I might be totally off base in this question BUT here goes...

In windows, if I ran into a problem that I couldn't figure out.. from corrupted drivers to well, basically any weird problem. I would totally uninstall networking and reinstall it. Is that a possibility or some idea like that just to start all over without wiping out the entire box?

try your kernel flag

/bin/echo "0" > /proc/sys/net/ipv4/icmp_echo_ignore_all


this turns icmp requests back on, also on redhat 6.2 you must likely are using ipchains for a firewall if you are using one at all, try doing an ipchains -L to see if you do have one running

Just in case I confused anyone, the Redhat 6.2 box is working fine, it's the Mandrake 9.2 box that I'm having trouble with.

Thanks!

OK, I feel like an idiot, sorry that I didn't know this. But my pings are working now.

The command to set my icmp_echo_ignore_all to "0" worked.

Thanks soo much everyone!

I'll probably be back in a little while, I think I still have another problem.

Thanks!!