RP5450 rebooted with few traces left

[Saurav_1]

Hi,

Please check the /var/adm/syslog/OLDsyslog.log

As root type 'elm' and say y for the next 2 prompts. you will get the mail in the folder.
If you find Event monitor Notification. Investigate them, Please attach them with your next post if you find anything interesting.

#set_fixed -L check the devices in down state. This may give some hint.

[mgais]

Saurav,
thanks for your hints but:
- NO messages logged in any syslog (I save syslogs even older then last OLDsyslog, in case a server reboots twice in a row)
- set_fixed says all devices are OK (and in fact unless the other day I would have noticed that a session on that server dropped, I wouldn't have noticed the reboot)

It seems the only trace, before a 'sudden death' that didn't give any time to write any log, are the MP logs posted yesterday.

I guess I have to open a call with HP to interprete them.

[Michael Steele_2]

When you have a HW problem, like you do now, the O/S has been taken out of the loop. So take the GSP logs and start a call with HW support. HW support will be able to interpret the codes for you. :-)

[Jeff Schussele]

Hi,

Most of the time when the ts99 contains no valid timestamps I've found that a bad CPU has panicked the system.
But HP can analyze the ts99 & tell you this for sure and also which CPU is the culprit.

HTH,
Jeff