Problem starting Hardware Monitoring

[Rajasekhar Raman]

I am having trouble Enabling Hardware Monitoring on my server. It was working just fine until a few days ago. I installed some security patches and ever since the reboot, the Hardware monitoring is not functioning. I have tried stop/starting "/sbin/init.d/diagnostic", and also moving the psm_data as was suggested in a thread, but to no avail.

I am able to start monconfig and set up the monitoring requests. When I check the detailed list of requests, I get a message

============================================================================
============== Check Outstanding Monitoring Requests ==============
============================================================================


Waiting for changes in monitoring requests or in hardware configuration
to take effect...

And it just hangs there. I checked the "/var/opt/resmon/log/client.log" file and I find the following messages for each resource monitor, looks like it is trying to get information by timing out:

-------------------Start Event--------------------
User event occurred at Thu Feb 24 14:23:29.428939 2005
Process ID: 10833 (/etc/opt/resmon/lbin/startmon_client) Log Level: Error
Resource name '/system/events/memory' timed out waiting for subclass reply

-------------------End Event----------------------

I have tried Killing the Monitoring and restarting but to no avail. This is a production system, so I cannot reboot anytime soon. All the diagnostic processes seem to be up, but none of the monitors seem to be running. I tried to manually startup the monitors in /usr/sbin/stm/uut/tools/monitors. To test whether everything was OK, I sent a "test event" and here is the response I got:

resls: Did not receive a Subclass Reply from the Registrar
/etc/opt/resmon/lbin/send_test_event: Failed to perform resls on resource name /storage/events/disks/default for disk_em to cause generation of test event.

Any help is appreciated.

-Shekar Raman

[Steven E. Protter]

Did these patches do the same thing when you installed into a test environment?

You have identified the most common cause of all problems, the last thing you did.

Two courses of action come to mind:

* Backing out the patches
* Installing additional patches to correct this problem.

You may need to open a case with the response center to figure out which is best.

I can see little alternative to scheduling a boot window on this box, unless you can live with this problem until your next scheduled window.

SEP

[Rajasekhar Raman]

I have installed the same patches on other servers without any problem. Is it possible that there is some corruption in one of the data files. Do you think uninstalling/reinstalling the Monitoring tools may clear it up?

-Shekar

[Robert-Jan Goossens]

Hi,

What is your current version of EMS ?

# /sbin/init.d/diagnostic stop

wait 60 seconds

# /sbin/init.d/diagnostic start

Regards,
Robert-Jan

[Rajasekhar Raman]

EMS-Config A.03.20 EMS Config
EMS-Core A.03.20 EMS Core Product
EMS-KRMonitor A.11.00.04 EMS Kernel Resource Monitor

I have tried to stopping and starting diagnostic as you suggested, and then Enabling Hardware Monitoring. Same result. When I enable thru monconfig, it hangs with the following message:
============================================================================
======================= Enable Monitoring ========================
============================================================================


This may take a while...

Waiting for changes in monitoring requests or in hardware configuration
to take effect...

and it seems like it is cycling thru all the monitors and not able to get a reply from them. Also at this point, I do not see any of the monitors up (like disk_em, sysstat_em etc.)

-Shekar

[Robert-Jan Goossens]

Hi,

Download the latest version from

http://www.software.hp.com/portal/swdepot/displayProductInfo.do?productNumber=B6191AAE

HP-UX 11.00 (Mar 04 release).
ID: B.11.00.27.xx, STM vers.: A.44.00

Check if your server needs other recommended patches BEFORE you update your EMS version.

http://docs.hp.com/en/diag/st/str_0403.htm

Regards,
Robert-Jan

[RAC_1]

ps -ef|grep diag

Is diaglogd and diagmond running?? Also after diagnostice start, do init q

Anil

[Andrew Merritt_2]

Did the patches you installed require a reboot, or did they change the run-level at all? What run-level is the system at now? Check with the ems* lines in /etc/inittab to see if the EMS daemons will be running at the run-level your system is at.

Did the patches have anything to do with networking?

Andrew

[Rajasekhar Raman]

Thanks for your responses, here are the answers to the questions posed.

1. Yes dialogd and diagmond are running
2. The patches did require a reboot, but so did the other 10 servers that I installed them on and which don't have a problem.

I will try the other suggestions and let you know. The last resort is to maybe remove the STM and EMS and re-install in the maintainence window.

-Shekar Raman