Hang system

[Pedro José]

We have an old HP 9000 D class.
It had installed an HP-UX 10 on it. One month ago, we installed HP-UX 11.
Till here, everything is OK.
Now, some days ago, the system is hanging without apparent reason.
When I realized the system has just hang I have a look at the display and it passes through the following states:
- RUN F71F
- RUN F91F
- RUN FA1F
When it comes to this final state, the system does not respond to any remote (telnet) nor console command. It only responds to a ping command.
What can be happening?
Thanks in advanced!

[Bernhard Mueller]

Pedro,

when you installed it, did you install the latest Patches (i.e. the QualityPack from the Support Plus CD)?

The behaviour of your system may be due to a runaway process or to a memory leak.

1st thing to do: get up to date with patches.

Regards
Bernhard

[T G Manikandan]

The best place to check is

/var/adm/syslog/syslog.log

[Stefan Farrelly]

Usually if you can ping it but not login then it has run out of memory to such and extent that it can no longer service login attempts.

Things to do are;
1. reboot
2. watch memory usage closely. Keep watching it to ensure its not increasing to such a point where you run out of memory (swapping starts). This is to be avoided. If it happens find out who is using the most memory and stop it, or add more RAM. Typically HP-UX 11.0 uses more memory than HP-UX 10.
3. Install glance/gpm to monitor memory usage closely and use this command to see which processes are using the most;
UNIX95= ps -e -o vsz=Kbytes -o ruser -o pid,args=Command-Line | sort -rnk1 | more

[Elmar P. Kolkman · Accepted solution]

One other thing: leave your console on. And log syslog messages to the console too (configurable in /etc/syslog.conf). That way, when the system hangs, you can look at the console for the latest messages. If you reset the machine, you might loose the latest messages in your syslog files, so this way you can view them before you reboot.

[Fabio Ettore]

Hi Pedro,

check these threads similar to your problem:

http://forums1.itrc.hp.com/service/forums/questionanswer.do?threadId=112625

http://forums1.itrc.hp.com/service/forums/questionanswer.do?threadId=97042

I hope this helps you.

Best regards,
Ettore

[doug hosking]

Fx1F, where 'x' is high means you have a 1-procesor system that is very busy. The higher 'x' is (up to a maximum of 'A') the busier the system is. There is not enough information in that display to tell much more than that. It could be something that is spawning many processes or any of a long list of other causes. Generally I would not suspect hardware as the most likely cause of this, but rather some software related cause. It could be a bug or something as simple as trying to run too much on a small system. Tools such as glance, started BEFORE there is a problem, may help you identify the real cause.
Alternatively, you could force a crash dump when the machine gets into that state and ask your HP support folks to take a look. (This type of problem is usually not hard to identify in a crash dump.)

If this happened to me I would try to login on the system and be patient. Chances are good that you will be able to login eventually, even if response time is very slow. From there you could do a lot more exploring than is possible just by looking at the front panel.

[kamal_9]

HI Jose
Go through the syslog and enable in syslog.conf to log the messages to the console also .it seems some process is using the full system.use dmesg and check wheather the system is showing the full memory .have a look on swap also.