topic Re: Server hung?? in Operating System - HP-UX

Server hung??

Tom Spence_1 — Thu, 28 Aug 2003 10:38:07 GMT

HP 877 10.20

The server always hung after I access to the server for 5 to 10 minutes. I kept reboot after it hung about 25 to 30 times a day. I don't see any problem, like what I did check: syslog.log (look good so far) and I checked all hard drives (48) by using dd command and all hard drives are good. What else should I check?

Re: Server hung??

Steven E. Protter — Thu, 28 Aug 2003 10:42:59 GMT

Look to recent changes in networking/nfs in /etc/rc.config.d/netconf

That is a common cause of hangs.

Also ping from another box. If the times are slowly rising, replace the NIC card. Watch it for an hour or so.

Run xstm or mstm and exercize the processor/disks/memory, anything that will excercize, you may find a problem.

SEP

Re: Server hung??

Chris Wilshaw — Thu, 28 Aug 2003 10:46:57 GMT

It could be a network issue.

Have you tried logging on from the console?

If that works OK, you need to have your network guys check things out.

Re: Server hung??

Tim Adamson_1 — Thu, 28 Aug 2003 10:49:42 GMT

Can you ping the server?
Can you telnet in from another server?
Can you login on the console?

Make sure you check OLDsyslog.log after the reboot as that is where any error messages would be if there were to be any.

Ensure patches are current, and ensure Online Diagnostics is installedand check the relevant log files (event.log). Obviously be careful installing any new patches or software as a hang 10 mins after logging in is not too good.

Tim

Re: Server hung??

Tom Spence_1 — Thu, 28 Aug 2003 10:50:04 GMT

I am using console all the time. It always hung after I access to the server for 5 to 10 minutes. I did ping from other box and the ping showed me which looks to me...

We don't use NFS... I check with NIC card is fine also.

Is this strange?

Re: Server hung??

Tom Spence_1 — Thu, 28 Aug 2003 10:53:43 GMT

Can you ping the server? YES
Can you telnet in from another server? YES
Can you login on the console? YES

Both console and telnet got hanging after 5 to 10 minutes too.

I have no problem with the server for 6 months after I patched. I did reboot once a week with no problem. After 6 months, it hung. Whoa... Not good...

Re: Server hung??

Tom Spence_1 — Thu, 28 Aug 2003 10:54:48 GMT

Oh, also OLDsyslog.log, it looks fine to me also... I guess the server is old (11 years old)!

Re: Server hung??

usergroup.os.ux-3.4-iro — Thu, 28 Aug 2003 10:58:28 GMT

could it be that a nameserver is not avail at this moment?
what is in your resolv.conf?

Re: Server hung??

Tom Spence_1 — Thu, 28 Aug 2003 11:29:39 GMT

The resolv.conf looks perfect...

Re: Server hung??

Massimo Bianchi — Thu, 28 Aug 2003 11:33:26 GMT

Could it be a memory leak somewhere ?

check, with cron

swapinfo -mt

every 5 minutes and log the results.

And, what about latest QPK ? how old is yours ?

Massimo

Re: Server hung??

Tom Spence_1 — Thu, 28 Aug 2003 11:49:15 GMT

It is 11 years old...

I will try swapinfo -mt after reboot...

I can ping from other box to the server which is great while server hangs now! Make sense to you?

Re: Server hung??

Kent Ostby — Thu, 28 Aug 2003 12:29:25 GMT

Tom --

In general when your system hangs, you can do a cntl-B> TC and force a crash dump of the box.

If you log a call with HP, they should be able to use that crash dump to troubleshoot the dump and recommend a course of action to repair it.

This document shows you how to pre-process your dump for HP:

http://www1.itrc.hp.com/service/cki/docDisplay.do?docLocale=en_US&docId=200000069402152

If you run the "whathappened" script that is mentioned, you ***MIGHT*** be able to see whats going on based on issues like what processes are running at the time of the hang or other data, but generally you need to get an HP support person involved.

Best regards,

Kent M. Ostby

Re: Server hung??

merydine — Fri, 29 Aug 2003 05:20:57 GMT

Hi evrybody,
I suggest "netstat -i" to find collisions.

J.M.Bourges

Re: Server hung??

Zigor Buruaga — Fri, 29 Aug 2003 05:48:24 GMT

Hi,

What happens if you don??t log on the machine, it also hungs?
What about booting in single user? or booting with the previous kernel? It also hungs ?

Maybe any error in /etc/rc.log or /etc/rc.log.old ( something starting in background that fails in 5-10 minutes? )
Only an idea.

Hope this helps.
Kind regards,
Zigor