Operating System - Tru64 Unix
1748237 Members
3608 Online
108759 Solutions
New Discussion юеВ

310 Tru64 UNIX Time Stamp Message

 
SOLVED
Go to solution
Chris Ellam
Advisor

310 Tru64 UNIX Time Stamp Message

Can anybody tell me what and when generates the binary log event 310 "Tru64 UNIX Time Stamp Message"? If possible a longer explination would be appreciated.

I am looking at 2 hung systems and 310 was the last event. There were no hardware errors and no messages/logs entries. I have no idea why the systems went away, there is no crash dump, there was no-reboots just a hang that required the operator to get in the SRM console and re-boot. Both systems, DS25s, hung over a weekend and needed a Monday morning re-boot.
4 REPLIES 4
Rob Leadbeater
Honored Contributor

Re: 310 Tru64 UNIX Time Stamp Message

Hi,

If I remember correctly, the timestamp messages are written to the binary event log periodically, simply so you can tell that the event logging system is working properly.

The fact that you have 310s up onto the hang, simply tells you that the logging system was working correctly up to that point...

I've seen DS25s exhibit similar problems when they get too hot. Do you have any environmental issues ?

Cheers,

Rob
Martin Moore
HPE Pro
Solution

Re: 310 Tru64 UNIX Time Stamp Message

To be more specific, the timestamp message gets added to the log when there has been no entry in the log for 10 minutes. If a timestamp was the last previous entry, the new timestamp replaces it, so you don't get a long string of timestamps with nothing else going on. As Rob says, this verifies that logging is functioning. It also gives you a rough time of when your hang actually occurred: sometime within 10 minutes after the last timestamp entry. If there was some other event that occurred in your environment around this time, perhaps you can find a correlation.

If you do find a hung system, it's worth trying to force a crash dump rather than just rebooting, since rebooting clears the contents of memory. The way to do this is to halt the system to get to the SRM console. How to halt the system depends on the system type -- on a DS25 and many other systems, there's a halt button on the front panel. Then when you get to the SRM prompt, enter "CRASH". The console firmware will dump the contents of memory to swap space, just as is done when the system panics. When the dump completes, boot the system normally. During the boot sequence, a startup script will detect the presence of the dump in swap space and save it to a crash dump in the crash directory -- /var/adm/crash by default, although you can change this in /etc/rc.config. The crash dump can be submitted to HP for analysis if you have Tru64 UNIX software support, or you can try looking at it yourself if you're knowledgeable (or adventurous).

Hope this helps,
Martin
I work for HPE
A quick resolution to technical issues for your HPE products is just a click away HPE Support Center
See Self Help Post for more details

Accept or Kudo

Chris Ellam
Advisor

Re: 310 Tru64 UNIX Time Stamp Message

Thanks Rob and Martin for the response and the clarification of what 310 does.

There is not an environmental issue, the systems are well cooled in the Air Traffic Controllers room.

Unfortunately the local system managers didn't make a forced crash which I'd liked to have a look at.

I suspected a memory error but on exercising the memory we got a SYSFAULT on CPU1 on 1 system.
Chris Ellam
Advisor

Re: 310 Tru64 UNIX Time Stamp Message

This has clarified what the 310 events do and mean. This gives an indication when the system went into its hung state and will be useful in associating this with the applications activity. The suggestion to force a crash dump when the system hangs will be implemented. Closing thread.