getting blank screen of death on RH8 ...

[K.C. Chan]

I am running rh8, kernel version 2.4.18-14smp, and 4Gbytes of memory. Does any one know why I am getting the blank screen of death; no response from server; I had to hard booted it. Does any one know how to get it to reboot if this event occures again? The log files is not very helpfull since the server locks up and I had to hard booted it.

[Mark Grant]

you need to be a bit more specific about whwn you get this. Is it only while running X (probably XF86Config or driver problem) in which case you could always try a CTRL-ALT-BACKSPACE when it happens. If it happens when no X is running it's a bit more serious and could be a number of things. You can always try CTRL-ALT-DEL to re-boot.

If all else fails you can always try the SysRq options which allow you to do things you really shouldn't be able to but might allow you to sync disks and re-boot in a crisis.

Check out this page

http://www.linuxhq.com/kernel/v2.4/10/Documentation/sysrq.txt

[Alexander Chuzhoy]

edit the /etc/sysctl.conf file
and change the kernel.sysrq = 0
to be
kernel.sysrq = 1
save the file and run: sysctl -p
after that if your server gets stuck you can try to execute alt+SysRq+K keys all together and it will log you out and close all running commands.
Sysrq is the Print Scrn Key.
Hope it'll help.
Best regards.

[Alexander Chuzhoy]

My previous message was regading command line. i.e. expiriencing "blank screen" and editing/executing file.

[Huc_1]

All I can think of is try and do a tail on all of the /var/log/* and zcat /var/log/*.1.gz if the log have been rotate and compress.

You could also do an xconsole & as root to a gui and let that run after having modify
/etc/syslogd.conf like

kern.* /dev/console
*.* /dev/console

Beware this will output lot of messages to console, you can tailor this for the types of messages you want

That way if the system stop you will have last console message frozzen on the gui screen.

You also need restart log services.

Hope this help.

J-P

[Mark Grant]

Are you really sure the server has locled up. That is very unusual in Linux. Youc an normally telnet to it if you have another machine to try it from.

How often does it happen and what graphics card to you have?

[K.C. Chan]

The server is not running X; just running console mode. A control+alt+delete is not responding either. Normally a control+alt+delete at the console reboots the server. The server is not ping-able either. I'm setting kernel.panic=10. I will let you know if it fixed the problem. This just starts happenning this last week and this week.

[Mark Grant]

Do you have power management enabled in the kernel? It might be worth trying to turn that off if you do.

[Alexander Chuzhoy]

I'm pretty sure that this is a hardware problem.Do you see some error messages in dmesg ?

[Huc_1]

Once again This is how I would tackel this

From an other system running X start an ssh session onto your problematic system
in that ssh session do

#su -

then do

#xconsole &

This will start a windows that will show all the messages you have set in /etc/syslog.conf
of your problematic system.

This will show you the last messages, I know you dont see much in the logs after the machine has crashed, but this will show you the sequence as it happens, and perhaps from this you can work it out.

J-P

[K.C. Chan]

how do I turned off pwr management, recompile the kernel or turned apmd off?

I've check dmesg, there are no error. I will try Huc's suggestion when I get into the office. Since I can't reproduce the problem, I will have to wait for the next event to occure. Thanks.

[Martin P.J. Zinser]

Have a look at the boot options disableapic and apm=off

Greetings, Martin