Downsample was originally written as a way to perform downsampling of time-series data in InfluxDB without adding additional load to the InfluxDB instance itself.
Its intended usage model was like that of Kapacitor, moving the CPU and RAM cost of aggregating data to the downsample node.
However, development ceased once the basic functionality was implemented and it's unlikely that there'll be much cause to pick it back up:
The current implementation does what I need of it so, rather than risking introducing new bugs, I'm ceasing development.
Activity