weird DHCP problems

weird DHCP problems

Post by Itay Kisho » Sun, 23 May 1999 04:00:00

Hello all.

I've been working with my RH5.2 installation for some time now,
connecting to the internet (through a cable modem) using an nic (NE2000
clone RealTek R-8029 PCI), with RedHat's automated setup for eth0 (All I
had to do is tell it to connect with DHCP, and list the DNS).

Yesterday I was configuring something in my linux, and as a result -
eth0 couldn't get the information from the dhcp server anymore. I tried
configuring many things in the system, and finally, after not being able
to get eth0 back to work, I reinstalled (!) RH5.2.

The weird thing is that even after a completly new installation, it
still doesn't aquire an IP from the dhcp server. I do know that the nic
and the driver are function correctly, because if I boot the system with
my Win98, write down the IP aquired from the dhcp, and then enter it
manually into eth0 configuration, the network operates find.

Also, when I try to run "dhcpcd -d" to aquire a new IP, I get the
following message: "bind (openSendSocket): Address already in use" (What
does it mean ?).

In spite all that, I did manage 2-3 times to get an IP from the dhcp,
after using netconf to restart the network. I haven't been able to do it
myself, using "ifconfig eth0 down"/"ifconfig eth0 up".

(A bit long, isn't it ? But I'm working on it for 2 days, without any

Any help would be appriciated,

        Itay Kishon.


1. weird DHCP problem

I am having a weird problem with DHCP. I am using a Linux box (RH7.1) connected
to a LAN. I have configured the Ethernet network card to use DHCP at boot time.
This is what happens in two different situations:

1.) the network cable from the Linux box is directly plugged into a LAN wall
when the machine is booting, it is not able to obtain a lease from the DHCP
server. I have tried to understand why, but I could not gather much
information. The DHCP server is working correctly, and when using a Windows box
I have no problems to obtain a lease. On the Linux client box, dhcpcd is
running. The logs (in /var/log) say that eth0 init failed but they did not
provide me any additional useful information.

Now this is what looks strange.
2.) the network cable from the Linux box is plugged into a switch and the
switch is plugged into a LAN wall socket: in this case, if I boot the machine
using the exact same configuration as above, the Linux client box obtains
successfully the lease from the DHCP server.

Can anyone please provide me any pointer to understand:

a.) where I can look for more detailed information about the cause of the
failure/why in case 1.) I fail to obtain a lease from the DHCP server.

b.) why it seems to make a difference using a switch.

Any suggestion is appreciated.
Thanks a lot. Max

Posted from []
via Mailgate.ORG Server - http://www.Mailgate.ORG

2. "Error Loading Ramdisk" during install

3. Weird DHCP problem on OpenBSD 3.2

4. adding man pages

5. Weird RH5.2/KDE/Qt/DHCP problem

6. Hacking, Security, Linux, Phreak, Exploits and more...

7. RH Linux 7 DHCP and weird network problem

8. advice needed .

9. Weird Ping, weird FTP, weird Telnet... HELP!!!

10. Weird, weird, weird issue ....

11. Weird DHCP occurence

12. Weird DHCP behavior

13. RH 5.1 flawless ftp-dhcp install; dhcp config retained; dhcp fails @ boot