How can I troubleshoot randomly rebooting 10.20 workstation?

[Jon Jenkins_2]

We have a workstation running 10.20 that has started rebooting randomly, starting once of twice a week, and now daily. I've checked /var/adm/syslog/syslog.log but there no entries, that starts just after the last reboot. What else can I check? Alternately how can I capture the error that is causing the reboot?

Many thanks.
Jon.

[T G Manikandan]

Did you check the OLDsyslog.log file.

That should have some hints.

It could be a memory overflow,CPU panic

REvert

[Jon Jenkins_2]

Thanks for the reply.

I have checked that file, but unfortunately the last entry is for Sep 16 12:58, and the reboot happened about 08:15 today (Sep 17) so there's no information to help.

[T G Manikandan]

Please do check /etc/shutdownlog
/etc/rc.log.old

for the shutdown information.

What is the workstation model?what are applications are running!

Revert

[T G Manikandan]

Do check

/var/crash
/var/tomstones dir's

Do you have any crash/dump files?

[Zigor Buruaga]

Hi,

If you don't find any related info in the suggested log files, maybe this colud help, little strange I know:
Several months ago we experienced something similar on a workstation. Sometimes it lost the /var filesystem, and hours after it goes down. Nothing special in syslog.log, nor shutdownlog, etc.
Since we lost one file system ( trying with mount -a reported that not recognized file system ), we suspected of disk status. All was apparently OK.
Out of ideas, we replace the disk and now all is OK. Only an idea.

Hope this helps.
Kind regards,
Zigor

[Paula J Frazer-Campbell]

HI

No entry in OLDsyslog.log tends to indicate a hardware shutdown as the OS did not have chance to write to syslog.

Have a look at current syslog for errors also dmesg.

HTH

Paula

[E. Wong]

Jon,

just a quick look at /etc/shutdownlog and /etc/rc.log can give u some hints.

check also for possible file systems that are full with bdf.

check with crontab -e for scheduled jobs that may be causing your wrkstn to reboot.

another suggestion, check for any files in /var/adm/crash. use strings command to read files.

Good Luck,

E.