HP UX 11i Log files to check

[John O'Neill_6]

Hi all,

A month or two ago we suffered a bad system crash, now I have a 'tail -f' going on the /var/adm/syslog/syslog.log file

I also run TOP, kmeminfo and sar regularly each day to check that the box is all ok.

So far so good, I've spoken to HP a few times and am told that system is ok.

Are there any other log files I should be checking on a regular basis for signs of hardware/operating system instability or issues?

I also regularly check the Chasis log.

I guess I just don't have the same faith in this box that I used to and need to know if what I am doing is enough to spot a potential hardware/OS failure before it happens.

Any tips appreciated.

-John

[A. Clay Stephenson]

You should really install the STM tools (Online Diagnostics); this toolset will continuously monitor your hardware. It would really help to know if your box is having hardware or OS problems. One of the most common causes of unreliable systems is failing to apply the periodic patch sets including the latest version of STM. The other thing to watch out for is excessive heat -- a good HVAC system and good, clean, reliable power is also a "must". Without knowing why your box crashed, it's diificult to be more specific. I assume you got HP to analyze your crash dump.

[Mel Burslan]

John,

In addition to what Clay has suggested, if you have a mission critical service support agreement with HP on this server, you may also consider installing HP's ISEE product and have HP monitor it around the clock for hardware and to some extent software failures. It is like the "ET Phone Home" kind of monitoring. It is mainly dependent on STM but instead of you babysitting the STM alerts, HP does the monitoring by sending alarms via http requests to their ISEE support center in Atlanta, which in turn initiate the support service calls in case some piece of hardware starts to show signs of malignancy.

I am not exactly sure what pricing structure ISEE entails but we have an enterprise-wide support contract with HP and ISEE monitoring is included in the package. You may want to check with your your ASE. I found it much better compared to getting paged for a failed root disk mirror in the middle of the night.

[Sudeesh]

STM is a great tool for sysadmin. It really help us to get alerted on any hardware issues from the 400 odd machines we are managing.

In addition keep watching /var/adm/crash and /var/tombstones/ts99.

Sudeesh

[morganelan]

You can use these tools for forcast your system:
1.HP Support Tool Manager (stm)
-Graphical (xstm) for X-based terminals.
-Menu (mstm), for HP ASCII terminals.
-Command Line (cstm) for all ASCII terminals.
2.EMS Hardware Monitors :
/etc/opt/resmon/lbin/monconfig
3.Check these log files:
/var/adm/syslog/syslog.log
/var/adm/crash/....0/*
/var/opt/resmon/logs/event.log
/var/tombstones/ts99

[Mahesh Kumar Malik]

Hi John

You may also monitor following log files as well regularly

/var/opt/resmon/logs/event.log
/var/amd/syslog/mail.log

Besides above /var/adm/wtmp file should be monitored for growth in size. If it grows, you may zero it down

Regards
Mahesh

[John O'Neill_6]

:)