disk threshold falling alert not generated, not reset

[tmagee]

Disk threshold falling alerts are not generated on a server I'm monitoring. Disk threshold rising alerts are generated but only one time either when disk space reaches the warning or critical setting, or if the settings applied are less than the current disk space available.

I'm running SIM 4.1, Windows 2000, monitoring the server that has SIM installed on it and get the same issue on management agents version 7.0 and 7.1.

I set up disk thresholds via SIM, configure, disk threshold, set disk thresholds, select the server I'm monitoring (a Proliant ML350G3) and use : -

critical threshold = 20
critical reset = 17
warning threshold = 13
warning reset = 10
agent polling interval =120

I add files to the C drive (there's only one drive in the system) and get alerts ok as it passes the warning and critical thresholds, but removing the files down to 6% free does not generate a falling alert (no entry in Windows 2000 Microsoft event viewer either for falling, but is there for rising, once only). Adding the files again passed the warning and critical values produce no alert.

Appears to be that the reset values are not being used by the agents but what is going wrong?

Tried also using the "File System Space Used" option in the server homepage but same result (how does this page know the reset values as the triangles only show the threshold values and not the reset values?).

SIM working ok in all the other alerts I've seen (though not tried the server utilization and processors threshold alerts).

Thanks in advance for suggestions / fix for this issue

[Ted Jerzewski]

Did you ever get fix/resolve for this issue .
Ref:disk threshold falling alert not generated, not reset

[tmagee]

Hello, thanks for checking into the forum issues. No, I've not been able to get a solution to this issue. I've installed SIM 4.2, (still on Windows 2000), and Proliant Support Pack v7.2 which installs management agents v7.2, but still the same issue.

I've been in contact with one of the U.S. based support personnel (got through sim page contacts), who found the same issue on their system. They contacted the SIM engineer who said there was no issue and it works as intended.

The engineer suggested that it could be an issue with SNMP running on the server,
and suggested to use the snmputil utility to test SNMP. I've not tried this out. I'm not getting event messages in W2k log related to the threshold so my suspicions lie with the agents.

The use of the disk capacity bar display in sim still confuses me, as you can set the warning and critical alert triangles through this but it doesn't show where the reset values are.

Let me know if you get any solutions yourself on this.

[tmagee]

I've read that a fix for this issue is planned for agents v7.4 (though don't take that as official confirmation of the issue or the fix).

Don't know if the: -
"Tried also using the "File System Space Used" option in the server homepage but same result (how does this page know the reset values as the triangles only show the threshold values and not the reset values?)." issue will be part of that, but this may /may not be in the sim 5.0 release. Anyone had success using this method?

[Darrin Rawls]

Yep, I think the agents had a bug. :( 7.40 should be shipping on the web on 9/9 last I heard...

[tmagee]

Found an HP customer advisory (ES050811_CW01) released on 26/8/05 describing

"On any ProLiant server running the HP Management Agents Version 7.40 (or earlier), if the disk threshold is exceeded, the agents will log the event in the Microsoft Windows NT Event Log. However, if the disk threshold is exceeded more than once, successive events will not be logged.

As a workaround, restart the HP Management Agents after a disk threshold event has been logged:

A future release of the HP Foundation Agents will allow the agents to log all disk threshold events. This advisory will be updated when additional information becomes available".

Hopefully a fix will be released.

[Nathan_44]

Has there ever been a fix for this issue? I am running SIM 5.0 with 7.40b agents and still cannot get a Falling Alarm for % Disk Space usage.
I have tried every Event Handler and have not got the falling alarm to work. Everything else works great.

Thanks!

[tmagee]

I've checked the customer advisory for this issue (ES050811_CW01) but no updates to it. I'm installing psp 7.40a as part of a general update, but if you're still getting the issue I'd expect our systems to do the same.

[tmagee]

There looks to be a fix for this issue released as part of management agents 7.41. "Threshold alerts would not get resent when the total disk usage falls below the threshold value and then crosses it again".

These agents are not part of PSP 7.4 but available for download separately. I've not tried them out yet but great news if this issue is corrected.

It doesn't mention if the disk threshold gui (the triangle settings to show critical values set) has been altered to show what the reset threshold values are though.

[tmagee]

Things are looking better now as regards disk threshold alerts. I'm running sim5.0, agents 7.41 on a w2k server with one logical disk volume I'm monitoring. With threshold values as in my original posting above, I get warning and critical alerts as disk usage crosses the values.

However I don't get a falling alert as disk usage drops passed the reset values, but I get warning and critical alerts ok as disk usage increases again. This wasn't happening with earlier agents so that's good now. Perhaps I can manage without the falling alerts.

(There's a 'absSuppressFallingTrap' listed in the rising alerts, perhaps this is related)?

But most of my other systems have 3 logical volumes that I want to set alerts for. I can use the file space gui triangles to set thresholds (these move ok for each volume), but how do I set the reset values (to allow a retrigger when space is cleared then filled again)?

The 'set disk thresholds' pull down in sim works ok for a single logical volume. It informs that its sent to the device successfully for systems with single and multiple disk volumes, but the gui doesn't change on a system with 3 logical volumes.

How do I set potentially different thresholds and reset values for a system with more than one logical volume?