nworks 5.5 RC2

[coen-v]

I've just upgraded our nworks 5.5 RC1 to 5.5 RC2. The update went smoothly without any problems. After the discovery of our Virtual Infrastructure environment (10 x 3.5 esx hosts) we got an unhealthy power sensor state on all the hosts. I noticed it was the System Board 1 Power Meter which was exceeding the threshold. I searched the documentation to find out the default value for the "nworks VMware: ESX Host Hardware Sensor has breached threshold".

I found two values for this monitor:

nworks VMware: ESX Host Hardware Sensor has breached threshold VMSENSOR. POWER Watts/Amps/kWh 2000 n/a no
nworks VMware: ESX Host Hardware Sensor has breached threshold VMSENSOR. POWER Watts/Amps/kWh 20 n/a no

Can somebody explain why there are two values in the documentation?
The default value in the Management Pack is 20 Watt, which is in my opinion very low. Maybe it's a small bug in the MP.

Greetings,

Coen

[Alec King]

Hi Coen,

Thanks! We also found this issue, unfortunately not in time for the RC2 build - but it will be fixed in our GA release.

The problem is caused because (in some servers) the POWER type sensor can produce metrics both in Amps, OR in Watts.

We do have two monitors - as you saw. One has threshold for 20 (amps) and one has threshold for 2000 (watts).

However in the RC2 build the monitors do not check for the metric unit type.

In our GA release we will include filtering for the monitors, so that the appropriate threshold runs against the appropriate monitor type.

For the moment you could work around this - for example, create two groups for POWER sensors, one where 'baseUnits' property = Watts and one where baseUnits=Amps. Then create overrides for the Hardware Sensor monitors that run against those groups.

And this problem is already fixed for our GA release - which is coming very soon!

Thanks again, and please let us know of any more feedback on MP RC2!

Regards
Alec

[coen-v]

Thanks for the fast reply. We will wait for the final release.

Greets,

Coen