Hi, everyone.
I know kernel oops can be seen by run 'dmesg', but if
kernel crashed, we can not run it. so I reconfigure syslogd
to support remote forward, the debug machine content of
syslogd.conf is:
##################
kern.* @192.168.28.137
(more lines after it are ignored)
##################
and run syslogd with '-m 0 -h' option.
the macheine have IP 192.168.28.137, its syslogd.conf:
##################
#kern.* /var/messages
(more lines after it are ignored)
##################
and I run syslogd on this machine with '-r' option.
After all, I run "tail -f /var/messages" on 192.168.28.137,
I can see boot log and normal printk() result. Well! however,
the most importantest message, crash Oops is lost.
Any suggest on it?
Wait for any reply. thanks in advanced.
sailor
On 9/2/05, liyu@WAN <[email protected]> wrote:
> Hi, everyone.
>
> I know kernel oops can be seen by run 'dmesg', but if
> kernel crashed, we can not run it. so I reconfigure syslogd
> to support remote forward, the debug machine content of
> syslogd.conf is:
When the panic is called, the network system cannt working, no
message will be sent. The panic is only designed to print at least
oops message on the screen.
For debug through ethernet, I suggest you to try KGDB, which consist
a patch to debug over ethernet.
--
Yingchao Zhou
On 9/2/05, liyu@WAN <[email protected]> wrote:
> Hi, everyone.
>
> I know kernel oops can be seen by run 'dmesg', but if
> kernel crashed, we can not run it. so I reconfigure syslogd
> to support remote forward, the debug machine content of
When the kernel crashes there's no guarantee that messages will reach
syslog. Actually there's no guarantee of anything - the kernel is
dead.
If you want to capture Oops messages in a more reliable fashion, then
use a serial console, netconsole or console on line-printer.
--
Jesper Juhl <[email protected]>
Don't top-post http://www.catb.org/~esr/jargon/html/T/top-post.html
Plain text mails only, please http://www.expita.com/nomime.html