Operating System - Tru64 Unix
1748232 Members
3364 Online
108759 Solutions
New Discussion юеВ

Re: Alpha Server gs160 hangs

 
Shirish Gadkari
Occasional Contributor

Alpha Server gs160 hangs

We have Alpha Server GS160 with O/s 5.1B. with 12GB memory and 8 CPus. We running Oracle 9i and the database is in archive log mode.

This is second time the server was hung and all the jobs cancelled. We had no any other alternative but to "boot" the system.

How can I find out what causes sytem to hang and how can fix this problem.

I need help ASAP because this is our production machine.

Thanks
4 REPLIES 4
Ivan Ferreira
Honored Contributor

Re: Alpha Server gs160 hangs

Install decevent and also WEBES. WEBES is a valuable tool (wsea command) that will help you to identify problems.

If the system did not generate a system crash, use the console crash command to force a memmory dump.
Por que hacerlo dificil si es posible hacerlo facil? - Why do it the hard way, when you can do it the easy way?
Shirish Gadkari
Occasional Contributor

Re: Alpha Server gs160 hangs

Thanks for the reply.
One more question. We do not have decevent on Tru64. You mentioned webes.. Can I down load window base and install on win XP to monitor Tru64.

What is easy way. I am not really a tru64 guru that is I would comfortable installing on windows XP but if I need to be on GS160 I can go thru trouble.
Thanks
Ivan Ferreira
Honored Contributor

Re: Alpha Server gs160 hangs

No, you can't. WEBES is tru64 specific.
Por que hacerlo dificil si es posible hacerlo facil? - Why do it the hard way, when you can do it the easy way?
Hein van den Heuvel
Honored Contributor

Re: Alpha Server gs160 hangs

>>> Alpha Server gs160 hangs
Question We have Alpha Server GS160 with O/s 5.1B. with 12GB memory and 8 CPus. We running Oracle 9i and the database is in archive log mode.

It used to work right?
So what were the latests changes (quantity and quality)
- more load?
- patches?
- tuning?

WAG ... you are in lazy (default) swap mode and have over commited memory and swap space.
Check with swapon -s once hte system is back up and running.

>> We had no any other alternative but to "boot" the system.

Call support. Involve more able people.

>> I need help ASAP because this is our production machine.

Call support. You are paying for that no?

>> How can I find out what causes sytem to hang and how can fix this problem.

Check log files: /var/adm/messages
Run (performance) monitors and most importantly actually try to interpret what they indicate: sar? collect? vmstat 100 100

>>> What is easy way. I am not really a tru64 guru that is I would comfortable

Call support. Warn your management that they gave you a task for which you are currently not properly equiped. Warn them politely that THEY, not you, are putting production at risk.

Hope this helps some,
Hein van den Heuvel (at gmail dot com)
HvdH Performance Consulting