This might be a bug and if not could someone explain to me how it works?
I have numerous windows CPU load sensors, Primary Channel is Total Percentage.
When i edit the channel and select "Enable alerting based on Limits" and then put a number inside the limits percentage and apply, The scale of the dial for the channel doesn't look right.
Example
Before.
After changing upper error limit.
Not sure why this would change, or how 104% is even possible.
Article Comments
Hi, Thanks for the quick reply.
I understand your example and it makes sense for measurements without hard boundaries however 100% is a percentage and given that CPU resource is a finite resource There can be no such thing as 120% CPU usage.
If a user sets an error state at 99% then I wouldn't expect to see anything but a small slither on the dial or a disproportionate scaling on the dial. If a user sets a error at 99% then when the measurement breaches that marker i would expect an alert but at the same when the measurement reached 100% then that's its limit.. 120% in this example means 100% so it misleading to anyone reading the data.
I understand that you are using the same dial and logic for percentage and infinite measurements but this does appear to be poor development if customers can end up with a figure above whats actually possible.
Apr, 2019 - Permalink
Hi there,
I surely understand your point, but for the gauge it is not really important what kind of sensor it is or what channel. There are currently no plans to change this behavior within PRTG.
Best regards.
Apr, 2019 - Permalink
Thanks, can you help me understand then why issues like the below image occurs. The settings are set to warn at 80 and error at 90...
Why does my machine goto 122% exactly?
If the difference between warning(80) and error (90) is 10% aka represented by the yellow then why is the red so big and why does it appear to have grown even more so since my original image.
Thanks
Apr, 2019 - Permalink
Hi there,
Could you check the "Maximum"-column of that channel? I suspect that it went above 100% one time. If this is the case, then please note that these values come directly from the device itself.
Best regards.
Apr, 2019 - Permalink
Hi you are correct, The column did go above 100% so just to be clear are percentage values also just read from the device? Ie there is no computational work done by PRTG to get the percentage value?
Kind Regards
Apr, 2019 - Permalink
Hi there,
The value is 1 to 1 taken from the device. This is a common issue, especially under Windows when the target device is totally overloaded.
Best regards.
Apr, 2019 - Permalink
Hi there,
When you add an error limit to a channel, then PRTG always add some percent to gauge in order to display this area. Example:
You have a channel with a maximum value of 50. The gauge will go from 0-50. You add a limit of 50 to the channel, now you wouldn't see any error area in the gauge. This is why PRTG adds a certain amount to make it visible.
Best regards.
Apr, 2019 - Permalink