Red Hat Bugzilla – Bug 58205
'ip link set eth1 up' hangs at boot time
Last modified: 2015-03-04 20:10:04 EST
From Bugzilla Helper:
User-Agent: Mozilla/4.78 [en] (X11; U; Linux 2.4.7-10 i686)
Description of problem:
-I can't get either of my two ethernet interfaces up at boot time:
The system hangs during boot when I do that.
-However I can start them manually after boot by using the 'ifup ethN' command.
-The tentative workaround of running say 'ifup eth1' from
/etc/rc.d/rc.local also halts the system.
-Since the system is halted I have to reset it, after which my filesystem
turns out corrupted: The file /var/lib/dhcpd/dhcpd.leases shows up as
an unattached inode and ends up in /lost+found after running fsck.
-I have been tracing the network scripts at boot time, finding out that
bringing up my eth0 card (BOOTPROTO=dhcp) makes the
system halt in the call '/sbin/dhcpcd -n -R eth0', whereas the eth1
card (no dhcp) halts the system during 'ip link set eth1 up'.
I presume the latter (ip) is the more basic executable. The -V option returns
'ip utility, iproute2-ss001007', and it came with rpm package iproute-2.2.4-14.
Actually I don't know if this is bug or if I'm doing something wrong.
Version-Release number of selected component (if applicable):
Steps to Reproduce:
1. Set ONBOOT=yes in /etc/sysconfig/network-scripts/ifcfg-eth1,
OR add line 'ifup eth0' in /etc/rc.d/rc.local
Actual Results: My system hangs every time at "Bringing up interface eth1"
Expected Results: It shouldn't have. It should have brought up eth1 and
I'm also running dhcpd and firewall rules at boot. Configuration file dhcpd.conf
of ipchains -L -v attached.
Created attachment 42253 [details]
Configuraion file of dhcpd
Created attachment 42254 [details]
output of 'ipchains -L -v'
OK, my first thought would be that something is broken with the initscripts. If
you could check the following things and/or provide the following information i
might be able to be of assistance:
1) Install the latest errata version of initscripts available via
http://www.redhat.com/support/errata/RHBA-2002-022.html. It fixes quite a few
network related problems.
2) If that doesn't help i need to know your hardware setup, mainly the exact
specs of your network cards. Additionally the rpm version numbers of the kernel,
initscripts, dhcp and initscripts are required.
Read ya, Phil
Hi Phil, thanks for your time. I ran up2date resulting in new
initscripts, but I still have the same problem. The package versions that
you asked for are currently the following:
(rpm -qa | grep kernel, initscript, dhcp)
As for the two ethernet cards:
3Com Eth.XL 3C905B-TX 10/100 NM
(Hope that makes any sense to you, I had these
names written in a file, I think I probably copied
them from the purchase bills).
The following might also be relevant:
1. I have a RH6.1 distribution deployed on another
disk partition of the same computer,
which doesn't have this problem.
2. I did some other (network-only) installations of the RH7.2 cd, still on the
same computer. They also suffer from the same problem,
but I a saw difference between two installations:
a. An intallation with high firewall security gave the
same problem as noted above.
b. An intallation with no firewall security behaved worse:
Ifup not only crashed at boot-time but also when
ran manually from a shell.
Note that all these different 7.2 installations were booted using
the same kernel: I.e., referring to disk partitions A, B and C,
A has a kernel, B and C are different installations.
LiloBoot1: bootkernel on A, /=B, /boot=A
LiloBoot2: bootkernel on A, /=C
(because C starts at high cylinder, I couldn't boot from C,
so I just reused A).
3. Also note that bringing up ethN has actually NEVER
worked correctly for me and 7.2:
I install 7.2-> I boot it-> It crashes -> I reset ->
I boot another partition -> I put ONBOOT=no in ifcfg-ethN ->
I boot -> It doesn't crash
4. By 'crash' I mean that no keystroke seems to have any effect,
including ctrl-alt-delete. Also I need only press my on/off switch
for a very short period of time to switch my computer off. Under
normal circumstances this is only the case after issuing the 'halt'
command. If I press my on/off button under normal circumstances
nothing happens unless I hold it for five seconds. So it would seem
that the crash moves the machine to a halted state.
5. Since the problem occurs with the two different cards, it might
be a motherboard thing: I have a Compaq presario 5476. But then
again, the 6.1 distribution has no problems.
I guess that's what I can think of at the moment,
hope there's a clue for you somewhere!,
Sorry for not responding earlier, but i was terribly swamped with work.
This really sounds very strange, especially that even after fresh installs the
problems persits and gets worse with no firewall rules set at all.
Could you do a
on that machine and attache the output here?
Also, did you give 7.3 a shot, especially the new kernel? It might be a driver
problem as it seems to really badly hang your machine...
Read ya, Phil
Created attachment 63266 [details]
Output of 'lspci -v'
Hi Phil, thanks for your time, I attached the output you asked for &
will upgrade to 7.3 in the next few days. I'll keep you posted,
Hi Phil, I gave up, 7.3 doesn't work either. I recently bought a new computer
and don't have any problems with that one. I'll give the compaq to my mother,
she's not very likely to post bugs here, so we can consider the bug closed!
Thanks for your time spent!,
OK, closing as worksforme (which it does. ;-).
Read ya, Phil