Reboot after panic

[Moon Sang Guk]

hpux 11.11 HP L2000
system reboot auto

shutdown log says...
->

15:58 Mon Dec 01 2003. Reboot after panic: SafetyTimer expired, isr.ior =
23:16 Mon Dec 08 2003. Reboot after panic: SafetyTimer expired, isr.ior = 0'10340004.0'dcf721d8
12:21 Wed Dec 10 2003. Reboot after panic: SafetyTimer expired, isr.ior = 0'0.0'0
~
what can i do.

[Steven E. Protter]

It would appear that something is possibly wrong with your processor.

You aren't real clear though. Is the system currently running? If so, you should look for crashes in /var/adm/crash

If their are subdirectories then there is a crash dump. Check the date.

Run q4 analysis on the dump and hp or someone smarter than me can tell you why the system crashed.

Attaching a a4 doc.

SEP

[KCS_1]

Hi,

At the first of all, you need to determine why the system crashed(either HPMC nor Panic)by 'q4' crash dump analizer.


#cd /var/adm/crash
#q4

q4> trace event 0

It may say "HPMC" or "Panic"

if,HPMC occured that's one of h/W problem

Call to HP or Service Provider.

[Moon Sang Guk]

q4> trace event 0
stack trace for event 0
crash event was a TOC
Send_Monarch_TOC+0x80
safety_time_check+0x170
per_spu_hardclock+0x4f8
clock_int+0x10c
mp_ext_interrupt+0x170
ivti_patch_to_nop3+0x0
idle+0x49c
swidle_exit+0x0


any problem ?

[melvyn burnard]

Looking at the info you have supplied, it appears you may be running Serviceguard, and you had a Serviceguard TOC. This means somehting has occurred to block the cmcld process from running.
Log a call with your local HP Response Centre and get the dump analysed properly.

[KCS_1]

Have a look at this below

http://www2.itrc.hp.com/service/iv/docDisplay.do?docId=/DE_SW_UX_swrec_EN_01_E/Dump.pdf


This will give you more the knoweldge about Dump on.

[Massimo Bianchi]

I agree with Melvyn, this looks like a SG issue.

The safety timer is used to reboot a node if it cannot talk to the other node.

Check in the syslog, you should find some trace of the kind: "turning safety timer on" or lost connection to cmcld daemon.

Was there any checnge to thi server recently, that can have caused this ?

Is this and heavy duty period, with network congestions, that can have caused this ?

Massimo

[Massimo Bianchi]

pardon,

OLDsyslog.log

Massimo