Conflict in throughput measure from built-in "Web Page Requests" BT
From my understanding, the throughput value in "Application Monitoring Overview" dashboard takes the count value of "Web Page Requests" BT, but I'm seeing a conflict in count values with maximum aggregation between "Application Monitoring Overview" dashboard and Custom chart.
"Application Monitoring Overview" dashboard:
Custom chart:
Both values takes value from count measure of "Web Page Requests" BT with maximum aggregation and are for the same application with same time frame and granularity(1 minute). I wonder why the value is so different(8670 in "Application Monitoring Overview", 13606 in custom chart)?
Hi
I got some more feedback from the engineering team - I hope the following statemente explains some of this data discrepancy between your two charts:
The charts with the baseline measurements are not one-minute resolution, but aggregated to a more coarse resolution.
This is because the “72 hours” timeframe is selected.
This aggregation causes some smoothing of the values where the actual maximum according to the underlying finest resolution gets lost.
The maximum which is displayed in the UI is just the actual maximum data point after aggregation, but not derived from the actual maximum of the underlying finest resolution.
Andi
Hi. I forwarded this to the engineering team that implemented these algorithms.
There are different algorithms we use for our baselining to "smooth" (not sure if this is the correct term) the data. It would explain a slight difference between charting the raw value and what we calculate for our baseline. Not sure if it explains that huge of a difference. I hope to get a more detailed answer from them and then post it here
Andi
