I’m developing a live streaming backend plugin with a historical data provider. The problem is: my historical backend does not guarantee that the historical datapoint result-count is equal to max-points. Instead it can happen, that I provide much more data in the requested time-frame. This isn’t a real problem for now. Grafana shows a compressed view I assume that some intermediate values are dropped during the rendering process, but to get an impression over the data everything is fine.
When I now go over to live-streaming the problem is, that grafana will drop all values which are larger than the max-datapoints options. So my result shown in the graf will not fill the selected timeframe, but is cutted after reaching max-points.
Wouldn’t it be a good Idea to use the max-point as rolling-buffer only when there is not a allocated buffer from a previously historical data request. And if the chart is already filled to use the current buffer to roll over instead? Or is there already an option which I missed?
Thank you for the feedback! I’ve forwarded it to the Live team.
Hi, any news on that?
It would be very helpful to see a solution for the rolling buffer topic. Should I create an issue in github to have a better tracking for that case?