Thanks for all the recommendations from:
Raul Sossa, Adrian Blount, Bryan Whiting, John Losey, George Gallen, Alan,
and Larry Clegg.
I like to post a respond from Bryan Whiting:
You might try checking /var/adm/syslog.dated and looking thru the various
log files. (looking at the directory that closely matches the time when the
system went down)
Also tail the /var/adm/messages file... you might see messages that indicate
something was failing prior to the crash.
If you have decevent installed, a quick look at "dia -R |pg" might help
too...
Worst case, if you have a support contract, log a call with the Support
Center, run a sys_check -escalate and submit the output for analysis... No
guarantees, but we might get lucky and find the cause...
Conclusion:
There is no event being recorded prior to the crash/reboot. I ran sys_check
-escalate to generate a group of logs and info files, but nothing shows
there was anything took place. I scheduled a downtime on Friday. Will
perform Alpha console/SRM level environment and crash analysis, such as
>>>show power. Hope can find the cause.
I suspect it is a electricity/power supply related incident. Years ago, I
had similar symptoms on OpenVMS systems on Alpha 8400 and VAX 7220. System
went down but no crash dump on OpenVMS. It was a case of sudden power
outage.
My original question is:
An Alpha 4100 hung. Could not access it from console either. I pushed the
"Reset' to force a reboot. The control panel's status window shows the
initial testings and then hangs on "Unload Console ...". Finally, I recycled
the power and got the 4100 reboot.
Log into the system, ran uerf, there is no event prior to the crash/reboot.
Other than the reboot, there is no record on the messages file either; could
not locate any core dump, and directory /usr/var/adm/crash is empty.
Thanks.
Warren
Received on Wed Apr 17 2002 - 22:16:13 NZST