why WAE entering kdb mode ?

[chia hao chang]

today....really not my day ><

in my user site environment

it's core WAE-7371 cms suddenly becomes offline and entering kdb mode

I never seen kdb mode and I find kdb mode from cisco

=====================

To enable access to the kernel debugger (kdb),  use the kernel kdb global configuration  command. Once enabled, kdb is automatically activated if kernel  problems occur, or you can manually activate it from the local console  for the WAFS device by pressing the required key sequence.

=====================

when it in kdb mode, the WAE HDD lights status is down, I cannot see any green light

I type reboot command to reboot WAE

but I didn't see any error or alarm in syslog

I've questions

1.why WAE to enter in kdb mode ?? is it the hardware issue or WAE kernel bug ??

2.entering kdb mode is cause kernel issue, why i cannot find error or alarm in syslog.txt

an appendix file is the syslog.txt

WAE-7371 cms offline in 2010/6/9 17:15 ~ 17:20 and I reboot WAE in 17:30

thanks

[Zach Seils]

The WAAS device enters kdb as the result of some type of catostrophic failure.  I would recommend you open a TAC case to proceed with troubleshooting.  If you'd rather we try to determine the root cause here on the forum, you'll need to provide the following information from when the WAE is in kdb:

At the "[?]kdb" prompt, enter the following commands. Notice how the  prompt changes when you change CPUs.

Enter at console                         Description

[?]kdb> bt                                  backtrace

[0]kdb> cpu 0                              change the CPU

[0]kdb> bt                                    backtrace (also  see below section on how to manually unwinding stack)

[0]kdb> rd                                   display register  values

[0]kdb> md %esp 16                      display memory

[0]kdb> md %ebp 16                   display memory

[0]kdb> sr 7                                 set the log level to  7

[0]kdb> cpu 1                            change the CPU, if  device has multiple CPU's

[1]kdb> bt                                 backtrace

[1]kdb> rd                                 display register  values

[1]kdb> md %esp 16                    display memory

[1]kdb> md %ebp 16                   display memory

[1]kdb> sr m                               show memory info

[1]kdb> set LINES 2500

[1]kdb> ps                               show processes

[1]kdb> dmesg                         display syslog  buffer

[1]kdb> md log_buf 2500         collect 2000 lines worth  of output which will show the unit status during boot-up

Note: when you see the "[0]more>"  or "[1]more>" just hit return, until you see the "[0]kdb>"  or "[1]kdb>" prompt again. Be  sure to CAPTURE ALL of the displayed information to a file.

You should also provide a sysreport from the device once it has been reloaded.

Thanks,

Zach