Solaris

I wonder if anyone here came accross the same situation like me.

One of server in my company reboot itself serveral days ago. I have checked /var/adm/messages... there is nothing there but a few messages mention explorer file has been generated by sysadmin. There is no core dump file either. SUN told me they cannot find root cause without core dump.

My question is that is there another to find root cause without core dump? Is there anything I might miss to check?

Thanks

My experiences have been with SCO only but I've seen the problem multiple times where the UPS battery has died and a power flicker was enough to cause a cold boot. No record on those systems other than the mail about the autoboot.

try the command "history"...to see if anybody rebooted it by mistake...or try to "more /.bash_history" or "more /.sh_history"...or maybe consider enabling auditing if you're suspecting one of your admins.


check your core dump configuration..?

type "dumpadm"...and see where your dump device is...why didn't it generate any core dump?


Check for any hardware failures...check if if the memory size is still the same...sometimes DIMMs fail and get disabled..but it shouldn't reboot. (# prtdiag -v)

Sometimes the board fail or memory DIMM fail..and it starts rebooting intermittently...and this will be an ugly situation.

If it was a genuine system reboot by itself,
provide me the model of the system, OS version, kernel patch info and OBP, prtdiag -v. Is it part of a cluster? If the system had rebooted during an explorer being run, then we will need to look into it as to why. Need your initial investigation input.

Thanks a lot, I will follow your advise and check it again today.