HP-UX - EMS to SIM via SNMP - very little information....

[David Libby_4]

Hello,

We're using the EMS wrapper to send EMS messages via SNMP to SIM. However, when SIM receives a message there is very little in the way of detail in said message. The alerts forwarded by mail from the CMS only contain criticity and high-level type.

This one, for example, is actually a power redundancy lost problem...

======================
Event Name: HP-UX EMS Critical Event
Event originator: hidden
Event Severity: Critical
Event received: 01-Apr-2006, 03:29:58
Event description: The Event Monitoring Service (EMS) has generated an event for a monitored resource. A Problem Event occurs when the threshold condition defined for the resource is true. A Change Event occurs when the monitored value changes. Monitoring requests configured for notification when the value changes may generate this event. A Poll Event occurs regularly at the configured polling interval. An Error Event occurs when the value of a monitored resource cannot be obtained from the monitor.
=======================

... not really very obvious that it has anything to do with power is it?

Clicking on the alert in SIM does give a tiny bit more detail but not the plain-English interpretations we are used with a mail directly sent from EMS itself.

This is annoying because it means we can't immediately determine what's going on and are forced not only to drill down... but to *login* to the machine involved and hunt in /var/opt/resmon/log/* for the details. This kind of defies the purpose of having centralised monitoring.

Does anyone else have this problem?

If this is only happening to me then I'll open a call with HP.

Best regards,
David Libby

[Andrew Merritt_2]

Hi David,
Have you looked at SysFaultMgmt? That integrates the EMS HW monitors with HP-SIM without using SNMP, and should give much better event reporting (you should see the actual event text on the central server).

http://www.hp.com/products1/unix/operating/SysFaultMgmt.html?jumpid=reg_R1002_USEN gives an overview, http://h20293.www2.hp.com/portal/swdepot/displayProductInfo.do?productNumber=SysFaultMgmt is the download page. (Due to a bug in ITRC, watch out for spurious characters particularly in that second URL.)

Currently the version available is for 11.23, but the 11.11 version will be released very soon (should be a matter of days).

Andrew

[David Libby_4]

Hi Andrew,

Question: can you please post a small example of what you see in a SIM event from EMS when SysFaultMgmt is installed correctly?

I have had a look at SysFaultMgmt and in fact *all* our machines have it installed 11.11 and 11.23. However we have some WBEM issues at the moment for which we have a call with HP open. :-/

SysFaultMgmt was not actually included on the HP-UX installation media until recently.

It seems that HP are finally getting round to bringing up to speed the monitoring on HP-UX/PA-RISC/IA, as compared to the Windows/Proliant side of things (which is *leaugues* ahead).

For now I'll keep an open-mind about it and will test as soon as our WBEM issue is resolved.

Regards,
David

[Andrew Merritt_2 · Accepted solution]

Hi David,
The improvement was made in the latest release, and I suspect if you've got SFM on 11.11 and 11.23 you will have the older version installed (11.11 support was withdrawn for a while, but will shortly be re-introduced).

Anyway, here's a section of the information you get when you click on the event in HP-SIM (assuming it formats ok here; it is much better laid out in HP-SIM. This is "Systems Insight Manager 5.0 with Update 1 - HP-UX", build C.05.00.01.00.06, and SFM B.01.00.10.02):

HP WBEM Event Details

IndicationIdentifier 103019700101000000
IndicationTime Tue, 3/28/2006, 2:44:27 PM WEST
Description This is a test message from the monitor to test the communication path from the monitor to the notification mechanism.
AlertingManagedElement
AlertType Device Alert
OtherAlertType
PerceivedSeverity Information
OtherSeverity
ProbableCause Unknown
ProbableCauseDescription <><> No action required.
Trending
EventID 103
EventTime Tue, 3/28/2006, 11:44:20 PM WEST
...


Andrew

[David Libby_4]

Hi Andrew,

Thank you very much for your informative feedback. Yes, you're right it's the old unsupported version which is installed.

Looks like I have a little updating of the CMS to do as well, as I am on SIM 5.0 (no update applied) migrated from 4.2.

The output you sent looks like it's going to be a *lot* more useful... looking forward to getting that functional.

Regards,
David

[David Libby_4]

Well it appears that this is not really a problem... but lack of functionality which is in the process of being changed by HP. A question of patience.