1748181 Members
4036 Online
108759 Solutions
New Discussion юеВ

Re: System Hangs

 
Michael LaRoche
Frequent Advisor

System Hangs

Hi,

I'm managing a system that hangs intermittently, it's running OpenVMS 6.2-1h3. It doesn't crash and we have to hit the reset button. The machine is an AlphaServer 400 4/233. It doesn't have most of the patches installed and I'm wondering if I need to install a patch and which one to fix the problem?

Thanks,
Mike
18 REPLIES 18
Kris Clippeleyr
Honored Contributor

Re: System Hangs

Mike,

Welcome.

Instead of hitting the reset button, can't you Ctrl/P on the console and then at the chevron prompt (>>>) type "crash".
This will produce a crashdump which can then be analyzed.

Greetz,

Kris (aka Qkcl)
I'm gonna hit the highway like a battering ram on a silver-black phantom bike...
Mohamed  K Ahmed
Trusted Contributor

Re: System Hangs

You should consider upgrading, there is a version 7.3-2 now
Also install the patches.
Most probably, the patches or upgrade will solve the problems

Mohamed
Ian Miller.
Honored Contributor

Re: System Hangs

anything in the error log?
check that you have the latest firmware
ftp://ftp.digital.com/pub/DEC/Alpha/firmware/archive/astn400.html
____________________
Purely Personal Opinion
Michael LaRoche
Frequent Advisor

Re: System Hangs

As for upgrading the system, that's not an option as the machine will be replaced within the next year.

We will try the ctrl/p the next time this happens.

Nothing in the error log for the time it hangs.
Robert Gezelter
Honored Contributor

Re: System Hangs

Mike,

Getting the crash is one thing. Applying the available patches for 6.21H3 and updating the firmware are good suggestions (at least they eliminate possibilities).

Can you tell us anything about what is happening on the system at the point that it apprears to hang? Is it hung for all users, or only for some users?

The more information that you can provide, the more helpful we can be.

- Bob Gezelter, http://www.rlgsc.com
Volker Halle
Honored Contributor

Re: System Hangs

Mike,

troubleshooting a hanging system is NOT trivial. Here are a couple of steps to take to at least get a forced crash, this is the ONLY way towards analysis of the hang,without pure speculation and guessing:

- the AlphaServer 400 does not have a separate HALT button. You need to re-jumper the Restart/Halt button to cause a HALT, so that you can type >>> CRASH on the console prompt, once the system is hung. CTRL-P may work on the serial console.

http://h18002.www1.hp.com/alphaserver/download/ek-pcdsa-ui-b01.pdf

- without a forced crash, errlog buffers from a hanging system can't be written to the system disk, so there is nearly no chance to find something in ERRLOG.SYS

Volker.
Wim Van den Wyngaert
Honored Contributor

Re: System Hangs

If you have performance advisor active, I would check high prio cpu loops, number of processes and memory usage before anything else.

Wim
Wim
Jan van den Ende
Honored Contributor

Re: System Hangs

Mike,


As for upgrading the system, that's not an option as the machine will be replaced within the next year.

Replaced by a newer VMS machine? In that case, it still is very wise to do the upgrade now, and then later the move.
That way, you separate any potential upgrade issues from any potential move issues.

-- any other replacement is to be considered a serious downgrade :-)

Proost.

Have one on me.

Seasonal greetings to all!

Jan
Don't rust yours pelled jacker to fine doll missed aches.
Anton van Ruitenbeek
Trusted Contributor

Re: System Hangs

Mike,


As for upgrading the system, that's not an option as the machine will be replaced within the next year.


An upgrade will aproaximaly take several hours. This is peanuts considering you got users who hang when the system hangs.
If you don't got a 24x7 company you can do this eq. in a weekend. If you got special applications the best thing is to hire for a week an Alpha, do an upgrade (not reinstal) of your current system on it and check. If everything is working OK, do it with you're production system. If you're in a cluster, you can do it on the fly .....

If you got a 24x7 environment, then you have off course a cluster and then you can test it also on the fly.

AvR
NL: Meten is weten, maar je moet weten hoe te meten! - UK: Measuremets is knowledge, but you need to know how to measure !