We have a single instance node, keeping all telemetry data in one single measurement. For each sensor we have at max 20 fields (float values). We are inserting 2.400.000 points per hour.
We want to save the daily data in an external csv file. This operation will be executed each day.
We’ll execute a query like that:
select * from telemetry where time >= 2017-09-01 00:00:00 and time <= 2017-09-01 23:59:59
What’s the impact of this kind of query on the Influx db server?
We saw a saturation of the RAM memory and then continuous swapping to the disk (the query takes really long time).
We have a single instance node, 4 cores, 8-10 GB RAM (the total RAM is 16GB but we have other processes running in it).