Hello,

The sensor of "Cloud HTTP v2", when viewed in the time window of "Live Data" shows as a numeric value of the HTTP response code (e.g. 200, 500, etc.), which is fine and shows well on the graph.

But, when the view is based on days, e.g. 2 days, 30 days or 365 days - it turns to be percent based, which I don't get the meaning of it this way.

I read the manual for this sensor but didn't find any explanation for this there.

Can someone explain this?

Article Comments

Hello,

By default this sensor is using a lookup file. So the "live Data" tab should show a string instead of the numerical value. Assuming the HTTP Response Code is 200, the live data-tab should show "OK". If it shows a numerical value instead, please check if a lookup file is selected in your channel settings.

On the 2 days, 30 days, and 365 days tabs on the other hand, PRTG always displays average values of the corresponding time span. If this average was an ordinary number, this would be easy. However, in case of the Cloud HTTP Sensor we are talking about a string ("OK") status and not a numeric value.

As an example: Let's assume your sensor showed "OK" twice and "Error" once within the last 2 days. PRTG would now need to create a average value - which is not really possible in this case without numeric numbers. Therefore PRTG uses the desiredValue attribute from the lookup file instead to calculate a percentage value.

In case of the Cloud HTTP sensor, the percent value in the 2 days graph of PRTG tells you how many percent of the time, sensor was "OK" (HTTP Response Code 200).


Mar, 2022 - Permalink

Thank you Timo!

1. Please, if you are a Peassler employee - please ask for someone to create and official and public KB for this, or elaborate about this in the sensor's details description article.

2. Let me see if I got this correctly for the sensor of "Cloud HTTP v2":

HTTP Response Code: Basically it is the numeric well-known HTTP response code (e.g. 200 = OK) In "Live Data" it shows on the graph the actual numeric value In X days sections it turns to be a percent (%) value and it simply shows a value of zero (0%) constantly, as if it does not calculate any trend of this value, as there is no meaning for doing so for numbers that represent very different things. So, since this attribute is not relevant in X days sections - better either remove it entirely in these sections, or show it but disable it and mark it as "N/A" and don't show it on the graph.

PRTG Cloud Response: A simple "On/Good/Working" or "Off/Bad/NotWorking" mark. In "Live Data" is shows either 1 (Working) or 0 (Not working) and in X days sections it shows these numeric values in percentage over time, between 100%, Working, to below that for non-working times

Is it so?


Mar, 2022 - Permalink

Hello,
1) This is also already explained in more detail here: https://helpdesk.paessler.com/en/support/solutions/articles/76000042282. This behavior applies to all sensors using strings.
2) Assuming the "Live-Data" shows "OK" permanently, the "x days" sections will show 100% for the corresponding channel. If the channel shows "Error" once in the "live-data"-tab, the percentage value in the "x days" section will drop, because it was no longer 100% ok for this time span. I'm afraid this behavior cannot be changed.


Mar, 2022 - Permalink