I have two boxes that were originally used as workstations at a school - they upgraded and sold these.

The two boxes are identical parts - Lenovo NetVista M42. I installed 5.3, installed RPMforge per the wiki instructions, did a yum -y update && shutdown -h now.

The next day I turned the machine on and installed the UT2004 dedicated server. After a few hours I tried logging in via SSH - I could enter my user and password, but could not get farther. I rebooted the server. After another day the same thing happened. I moved the disk from the first box to the second and moved /etc/sysconfig/network-scripts/ifcfg-eth0.bak to ifcfg-eth0 but retaining the correct MAC address.

During this transition where the server "locks up" some things work and others do not. For example, if I type ping 4.2.2.2, wait 10 minutes and CTRL+C it looks like this (note the time of 0ms):

# ping 4.2.2.2
PING 4.2.2.2 (4.2.2.2) 56(84) bytes of data.
64 bytes from 4.2.2.2: icmp_seq=1 ttl=54 time=32.1 ms

--- 4.2.2.2 ping statistics ---
1 packets transmitted, 1 received, 0% packet loss, time 0ms
rtt min/avg/max/mdev = 32.152/32.152/32.152/0.000 ms

watch -n1 date does not update

If I type the command date I do see the correct time.

I have had ntpd running and not running - the problem continues. There is no virtualization. dmesg > temp.txt caused my SSH session to halt and I had to close the window.

dmesg - http://pastebin.ca/1648898
lspci - http://pastebin.ca/1648901

I've never experienced any problem like this before so please let me know if I am missing anything that can help troubleshoot this issue.

Thanks in advance!

The only thing interesting I could find in your dmesg is the following:


* The chipset may have PM-Timer Bug. Due to workarounds for a bug,
* this clock source is slow. If you are sure your timer does not have
* this bug, please use "acpi_pm_good" to disable the workaround


Can you try adding acpi_pm_good to your kernel command line?

I've seen the above message quite a few times but never had problems with it before.

I made these changes to /boot/grub/menu.lst (menu.lst softlink)

title CentOS (2.6.18-164.el5)
root (hd0,0)
kernel /vmlinuz-2.6.18-164.el5 ro root=LABEL=/ acpi_pm_good=1
initrd /initrd-2.6.18-164.el5.img

Then issued command "reboot"

Is this what you asked me to do?

I changed the kernel options as I said above, but the server once again locked up.

I've been having the same issue ever since updating centos 5.3 to 5.4

Using custom compiled kernel that worked fine on 5.3 and now things randomly lock up on 5.4 exactly the way you describe.

Looks like something that changed in 5.4 is causing this at least for me.

Running CentOS release 5.3 (Final) kernel /vmlinuz-2.6.18-164.el5 ro root=LABEL=/ acpi_pm_good=1

Does anyone have any other ideas or suggestions that may lead to a solution?

Did it start doing this suddenly or has it always done it since the install?

Have you tried on an older or a newer kernel?

Try 'clocksource=pit' as an argument on your kernel command line?

Hi mate

I had exactly your same issue, a year ago I bought an old IBM netvista pIV 1,8 ghz in ebay to use it with elastix, that is a centos flavour for asterisk.
I had same message about timer bug in dmesg, after a lot of google and headache I found somewhere that those computers where made with a kind of bug with acpi and apm. I can't rememember exactly how I have grub in that server I think is: acpi=off and noapic in the end of kernel line.

Try that and tell us if it works

Quote:

Problem exists with the install kernel and the update kernel

title CentOS (2.6.18-164.el5)
root (hd0,0)
kernel /vmlinuz-2.6.18-164.el5 ro root=LABEL=/
initrd /initrd-2.6.18-164.el5.img
title CentOS (2.6.18-128.el5)
root (hd0,0)
kernel /vmlinuz-2.6.18-128.el5 ro root=LABEL=/ acpi_pm_good=1
initrd /initrd-2.6.18-128.el5.img


Quote:

Trying this tonight

title CentOS (2.6.18-164.el5)
root (hd0,0)
kernel /vmlinuz-2.6.18-164.el5 ro root=LABEL=/ acpi=off noacpi
initrd /initrd-2.6.18-164.el5.img

This seems to have done the trick. The server has been up for 36 hours now (while before it would "freeze" after as little as 6).

title CentOS (2.6.18-164.el5)
root (hd0,0)
kernel /vmlinuz-2.6.18-164.el5 ro root=LABEL=/ acpi=off noacpi
initrd /initrd-2.6.18-164.el5.img

By-the-way the instructions said to use "acpi_pm_good" NOT "acpi_pm_good=1".