Alarm Tolerance Not Working?

[rmancinelli]

Hello,

Under Alarm Management, Virtual Machine, Guest Disk Space, I have the alert configured as follows:

Rules Tab: Enabled, Type Relative, Condition Below, Warning % 10, Error % 5
Reporting Tab: Tolerance 20%, Frequency 60 minutes

I had originally thought that the tolernance was measured in the same way as the warning and error percentages are measured. I was corrected, and told that the Tolerance was actually RELATIVE to the Warning and Error percentages... thus my tolernance of 20% is really 20% of 10% of the warning level, which (as I understand) should mean that the alarm should not be repeated unless the disk free space leaves 8% range (10% - (20% of 10%)).

Regardless of how I set this, I get the same alert every 60 minutes. Is there a bug?

Rick

[Vitaliy S.]

Hi Rick,

You're almost right. 20% of Tolerance leaves you with 2% above and below the default value of 10%, in other words you have 4% range (not 8%) for the free space.

Here is a brief example:
Disk capacity - 100 GB
Tolerance - 20% (2 GB)
Warning threshold - 10% (10GB)

The alarm will be triggered if only you exceed 12-8 GB interval. Could you please check that the free space on your source disk meet 4% range requirements.

P.S. Please also note that the tolerance will only be used on the "second" check of disk free space, meaning that as soon as the free space on your disk reaches 10% of total disk capacity the alarm will be raised in any case. The second alarm will only be raised if the level of free disk space is out of the defined tolerance range.

Thanks,
Vitaliy

[rmancinelli]

Vitaly,

I understand that the range would be +/- 2%, which would give us from 12% to 8% as a range, or 4% in total. I figured the 12% portion was inconsequential because once we're over 10%, the alarm should stop anyway.

That said, the disk space is NOT fluttering +/- 2%. It is saying fairly constant and we do get these alarms every 60 minutes without fail. I have tried rebooting the Monitor server as well, but this didn't help.

I saw somewhere else that there was a hotfix associated with reading available diskspace. It did not appear to apply to this particular instance, but I thought I would raise that point as well.

Rick

[Vitaliy S.]

Rick, thanks for clarifications. Yes, there is a hotfix for guest disk space alarm triggering, but this hotfix doesn't relate to tolerance engine, though If I were I would have applied it anyway.

On top of that, if you still have alarms (triggered despite the tolerance range), please contact our techincal team for assistance directly.

P.S. Also please PM your ticket number.