Event Monitoring Services in SAM showing error for file system utilisation

[James Bell_1]

Hi,

I have a hpux 11.00 system and I monitor the disk utilisation of the system through "Event Monitoring Services" in SAM, under Resource Management.

All of the file systems that I monitor at the moment are listed as a value of either, "Error" or as "Unknown". They usually tell me what the utilisation of the file system is and then page me when they are getting full.

I can see all the disk space correctly through bdf, and I don't have any errors in syslog.log.

How do I go about getting the values reported again in SAM.

Any help would be greatly appreciated. Thanks.
James.

[Steven E. Protter]

Paste the error in here please.

I'm pasting in a disk utilization script you can run out of cron that doesn't involve EMS.

It can be modified to monitor any percentage criteria for notification. It can also send the results out via sendmail.

See the attachment.

SEP

[James Bell_1]

From the window Event Monitoring Service that is accessed from going through SAM -> Resource Management -> Event Monitoring Services. I get a window with this information.

Resource Value Notify
/system/filesystem/availMb/opt Unknown When value...
/system/filesystem/availMb/var Unknown When value...

There are more options in this window that don't fit here.

Normally, the disk space utilisation would be in the Value field and would be something such as 252.3000 - indicating I have this amount of available space.

Thanks
James

[Bhuvaneswari Selvaraj]

Hi,

Can you please check whether the SNMP daemon and the other agents are running on your system or not? If not,
can you please start them. You can do so using the following commands.

ps -ef | grep snmpdm
ps -ef | grep hp_unixagt
ps -ef | grep trapdestagt
ps -ef | grep mib2agt

If they are not running, start them as follows
/sbin/init.d/SnmpHpunix start
/sbin/init.d/SnmpMaster start
/sbin/init.d/SnmpTrpDst start
/sbin/init.d/SnmpMib2 start

Exit the GUI and restart the same by clicing on the "Resource Management" icon in SAM.

Please do let me know, whether this solved the problem or not.

[Bhuvaneswari Selvaraj]

Hi,

Can you please check whether the SNMP daemon and the other agents are running on your system or not? If not,
can you please start them. You can do so using the following commands.

ps -ef | grep snmpdm
ps -ef | grep hp_unixagt
ps -ef | grep trapdestagt
ps -ef | grep mib2agt

If they are not running, start them as follows
/sbin/init.d/SnmpMaster start
/sbin/init.d/SnmpMib2 start
/sbin/init.d/SnmpHpunix start
/sbin/init.d/SnmpTrpDst start

Exit the GUI and restart the same by clicing on the "Resource Management" icon in SAM.

Please do let me know, whether this solved the problem or not.

[Rainer von Bongartz]

I guess you're having an snmp problem on you machine.

take a look at

http://www4.itrc.hp.com/service/iv/docDisplay.do?docId=/DE_SW_UX_swrec_EN_01_E/EMS.pdf

to see what snmp requirements are necessary.


Regards
Rainer

[James Bell_1]

Hi all,

I have restarted all of the snmp components on the system as suggested and have progressed a little.

I now get a mixture of some of the values being displayed as "rubbish numbers", "unknown" or "error".

I have also gone and restarted ems with monconfig as suggested in the EMS section of the manual, but this has not had any effect on the values being displayed from when I restarted snmp.

This step of restarting snmp has obviously had some effect but not fully refreshed the values. There must still be some other area where the values are stored or processed that is still not working correctly yet.

Thanks very much for all the help so far and for any ideas about how to finally resolve this.

Thanks
James

[Bhuvaneswari Selvaraj]

Hi,

May be you can try this. Exit out of the GUI.

Do a
ps -ef | grep resmon
Kill all the processes that get listed (all the monitors, registrar and p_client)

Restart the GUI.

If this also does not work, i guess that we need to analyse the EMS logs under /etc/opt/resmon/log. Is the dir empty? If not, just see whether any of these logs show any error (grep Error *.log) and let us know the error.