This is a more of an open issue. We want to maximize efficiency of queries done to DB, especially because they are done on Raspberry Pi with limited computing power. Right now we ask db for points in 1 second intervals, so when we ask for 1hour of historic data it responds with 3600 points.
It was done that way because ON/OFF telemetry could be lost when larger intervals were to be used. But for normal temperature / pressure values I think we can make the interval larger and maybe use "mean" aggregation.
So overall, the task is to measure current time of query and try to lower it.
This is a more of an open issue. We want to maximize efficiency of queries done to DB, especially because they are done on Raspberry Pi with limited computing power. Right now we ask db for points in 1 second intervals, so when we ask for 1hour of historic data it responds with 3600 points.
It was done that way because ON/OFF telemetry could be lost when larger intervals were to be used. But for normal temperature / pressure values I think we can make the interval larger and maybe use "mean" aggregation.
So overall, the task is to measure current time of query and try to lower it.