Data aggregation other than "average"?



  • While an averaging over time periods is certainly useful, I’d also like to have an easy way to catch spikes / maximum usage.

    For instance, this is a system.net graph zoomed out:

    943c6801-777c-401e-b5ce-a035608b47b0-image.png

    And now I’ll zoom in on the spike at 08:ish:
    75f08fd7-3ae0-4740-8a1c-fdb493f0b952-image.png

    You can see that the maximum jumped from 6 mb/s to 40 mb/s! Which makes it rather hard to differentiate between a baseload and a throttled connection due to reaching a (bandwidth) cap.



  • We decided to start with just average, but we will be adding more numeric aggregation methods next.


Log in to reply